displaying: http://coolnspicy.com/feedMagpieRSS Object
(
[parser] => 0
[current_item] => Array
(
)
[items] => Array
(
[0] => Array
(
[title] => This simple material could scrub carbon dioxide from power plant smokestacks
[link] => https://coolnspicy.com/science/this-simple-material-could-scrub-carbon-dioxide-from-power-plant-smokestacks/
[dc] => Array
(
[creator] => Michael Steiner
)
[pubdate] => Thu, 03 Nov 2022 08:17:48 +0000
[category] => sciencecarbondioxidematerialplantpowerscrubsimplesmokestacks
[guid] => https://coolnspicy.com/?p=299
[description] => Journal Reference: Hayden A. Evans, Dinesh Mullangi, Zeyu Deng, Yuxiang Wang, Shing Bo Peh, Fengxia Wei, John Wang, Craig M. Brown, Dan Zhao, Pieremanuele Canepa, Anthony K. Cheetham. Aluminum formate, Al(HCOO) 3 An earth-abundant, scalable, and highly selective material for CO 2 capture. Science Advances, 2022; 8 (44) DOI: 10.1126/sciadv.ade1473 The team’s object of study ... Read more
The post This simple material could scrub carbon dioxide from power plant smokestacks first appeared on .
[content] => Array
(
[encoded] =>
Journal Reference:
- Hayden A. Evans, Dinesh Mullangi, Zeyu Deng, Yuxiang Wang, Shing Bo Peh, Fengxia Wei, John Wang, Craig M. Brown, Dan Zhao, Pieremanuele Canepa, Anthony K. Cheetham. Aluminum formate, Al(HCOO) 3 An earth-abundant, scalable, and highly selective material for CO 2 capture. Science Advances, 2022; 8 (44) DOI: 10.1126/sciadv.ade1473
The team’s object of study is aluminum formate, one of a class of substances called metal-organic frameworks (MOFs). As a group, MOFs have exhibited great potential for filtering and separating organic materials — often the various hydrocarbons in fossil fuels — from one another. Some MOFs have shown promise at refining natural gas or separating the octane components of gasoline; others might contribute to reducing the cost of plastics manufacturing or cheaply converting one substance to another. Their capacity to perform such separations comes from their inherently porous nature.
Aluminum formate, which the scientists refer to as ALF, has a talent for separating carbon dioxide (CO2) from the other gases that commonly fly out of the smokestacks of coal-fired power plants. It also lacks the shortcomings that other proposed carbon filtration materials have, said NIST’s Hayden Evans, one of the lead authors of the team’s research paper, published today in the peer-reviewed journal Science Advances.
“What makes this work exciting is that ALF performs really well relative to other high-performing CO2 adsorbents, but it rivals designer compounds in its simplicity, overall stability and ease of preparation,” said Evans, a chemist at the NIST Center for Neutron Research (NCNR). “It is made of two substances found easily and abundantly, so creating enough ALF to use widely should be possible at very low cost.”
The research team includes scientists from the National University of Singapore; Singapore’s Agency for Science, Technology and Research; the University of Delaware; and the University of California, Santa Barbara.
Coal-fired power plants account for roughly 30% of global CO2 emissions. Even as the world embraces other energy sources such as solar and wind power that do not generate greenhouse gases, finding a way to reduce the carbon output of existing plants could help mitigate their effects while they remain in operation.
Scrubbing the CO2 from flue gas before it reaches the atmosphere in the first place is a logical approach, but it has proved challenging to create an effective scrubber. The mixture of gases that flows up the smokestacks of coal-fired power plants is typically fairly hot, humid and corrosive — characteristics that have made it difficult to find an economical material that can do the job efficiently. Some other MOFs work well but are made of expensive materials; others are less costly in and of themselves but perform adequately only in dry conditions, requiring a “drying step” that reduces the gas humidity but raises the overall cost of the scrubbing process.
“Put it all together, you need some kind of wonder material,” Evans said. “Here, we’ve managed to tick every box except stability in very humid conditions. However, using ALF would be inexpensive enough that a drying step becomes a viable option.”
ALF is made from aluminum hydroxide and formic acid, two chemicals that are abundant and readily available on the market. It would cost less than a dollar per kilogram, Evans said, which is up to 100 times less expensive than other materials with similar performance. Low cost is important because carbon capture at a single plant could require up to tens of thousands of tons of filtration material. The amount needed for the entire world would be enormous.
On a microscopic scale, ALF resembles a three-dimensional wire cage with innumerable small holes. These holes are just large enough to allow CO2 molecules to enter and get trapped, but just small enough to exclude the slightly larger nitrogen molecules that make up the majority of flue gas. Neutron diffraction work at the NCNR showed the team how the individual cages in the material collect and fill with CO2, revealing that the gas molecules fit inside certain cages within ALF like a hand in a glove, Evans said.
Despite its potential, ALF is not ready for immediate use. Engineers would need to design a procedure to create ALF at large scales. A coal-fired plant would also need a compatible process to reduce the humidity of the flue gas before scrubbing it. Evans said that a great deal is already understood about how to address these issues, and that they would not make the cost of using ALF prohibitive.
What to do with the CO2 afterward is also a major question, he said, though this is a problem for all carbon-capture materials. There are research efforts underway to convert it to formic acid — which is not only a naturally occurring organic material but also one of the two constituents of ALF. The idea here is that ALF could become part of a cyclic process where ALF removes CO2 from the exhaust streams, and that captured CO2 is used to create more formic acid. This formic acid would then be used to make more ALF, further reducing the overall impact and cost of the material cycle.
“There is a great deal of research going on nowadays into the problem of what to do with all the captured CO2,” Evans said. “It seems possible that we could eventually use solar energy to split hydrogen from water, and then combine that hydrogen with the CO2 to make more formic acid. Combined with ALF, that’s a solution that would help the planet.”
This simple material could scrub carbon dioxide from power plant smokestacks
The post This simple material could scrub carbon dioxide from power plant smokestacks first appeared on .
)
[summary] => Journal Reference: Hayden A. Evans, Dinesh Mullangi, Zeyu Deng, Yuxiang Wang, Shing Bo Peh, Fengxia Wei, John Wang, Craig M. Brown, Dan Zhao, Pieremanuele Canepa, Anthony K. Cheetham. Aluminum formate, Al(HCOO) 3 An earth-abundant, scalable, and highly selective material for CO 2 capture. Science Advances, 2022; 8 (44) DOI: 10.1126/sciadv.ade1473 The team’s object of study ... Read more
The post This simple material could scrub carbon dioxide from power plant smokestacks first appeared on .
[atom_content] =>
Journal Reference:
- Hayden A. Evans, Dinesh Mullangi, Zeyu Deng, Yuxiang Wang, Shing Bo Peh, Fengxia Wei, John Wang, Craig M. Brown, Dan Zhao, Pieremanuele Canepa, Anthony K. Cheetham. Aluminum formate, Al(HCOO) 3 An earth-abundant, scalable, and highly selective material for CO 2 capture. Science Advances, 2022; 8 (44) DOI: 10.1126/sciadv.ade1473
The team’s object of study is aluminum formate, one of a class of substances called metal-organic frameworks (MOFs). As a group, MOFs have exhibited great potential for filtering and separating organic materials — often the various hydrocarbons in fossil fuels — from one another. Some MOFs have shown promise at refining natural gas or separating the octane components of gasoline; others might contribute to reducing the cost of plastics manufacturing or cheaply converting one substance to another. Their capacity to perform such separations comes from their inherently porous nature.
Aluminum formate, which the scientists refer to as ALF, has a talent for separating carbon dioxide (CO2) from the other gases that commonly fly out of the smokestacks of coal-fired power plants. It also lacks the shortcomings that other proposed carbon filtration materials have, said NIST’s Hayden Evans, one of the lead authors of the team’s research paper, published today in the peer-reviewed journal Science Advances.
“What makes this work exciting is that ALF performs really well relative to other high-performing CO2 adsorbents, but it rivals designer compounds in its simplicity, overall stability and ease of preparation,” said Evans, a chemist at the NIST Center for Neutron Research (NCNR). “It is made of two substances found easily and abundantly, so creating enough ALF to use widely should be possible at very low cost.”
The research team includes scientists from the National University of Singapore; Singapore’s Agency for Science, Technology and Research; the University of Delaware; and the University of California, Santa Barbara.
Coal-fired power plants account for roughly 30% of global CO2 emissions. Even as the world embraces other energy sources such as solar and wind power that do not generate greenhouse gases, finding a way to reduce the carbon output of existing plants could help mitigate their effects while they remain in operation.
Scrubbing the CO2 from flue gas before it reaches the atmosphere in the first place is a logical approach, but it has proved challenging to create an effective scrubber. The mixture of gases that flows up the smokestacks of coal-fired power plants is typically fairly hot, humid and corrosive — characteristics that have made it difficult to find an economical material that can do the job efficiently. Some other MOFs work well but are made of expensive materials; others are less costly in and of themselves but perform adequately only in dry conditions, requiring a “drying step” that reduces the gas humidity but raises the overall cost of the scrubbing process.
“Put it all together, you need some kind of wonder material,” Evans said. “Here, we’ve managed to tick every box except stability in very humid conditions. However, using ALF would be inexpensive enough that a drying step becomes a viable option.”
ALF is made from aluminum hydroxide and formic acid, two chemicals that are abundant and readily available on the market. It would cost less than a dollar per kilogram, Evans said, which is up to 100 times less expensive than other materials with similar performance. Low cost is important because carbon capture at a single plant could require up to tens of thousands of tons of filtration material. The amount needed for the entire world would be enormous.
On a microscopic scale, ALF resembles a three-dimensional wire cage with innumerable small holes. These holes are just large enough to allow CO2 molecules to enter and get trapped, but just small enough to exclude the slightly larger nitrogen molecules that make up the majority of flue gas. Neutron diffraction work at the NCNR showed the team how the individual cages in the material collect and fill with CO2, revealing that the gas molecules fit inside certain cages within ALF like a hand in a glove, Evans said.
Despite its potential, ALF is not ready for immediate use. Engineers would need to design a procedure to create ALF at large scales. A coal-fired plant would also need a compatible process to reduce the humidity of the flue gas before scrubbing it. Evans said that a great deal is already understood about how to address these issues, and that they would not make the cost of using ALF prohibitive.
What to do with the CO2 afterward is also a major question, he said, though this is a problem for all carbon-capture materials. There are research efforts underway to convert it to formic acid — which is not only a naturally occurring organic material but also one of the two constituents of ALF. The idea here is that ALF could become part of a cyclic process where ALF removes CO2 from the exhaust streams, and that captured CO2 is used to create more formic acid. This formic acid would then be used to make more ALF, further reducing the overall impact and cost of the material cycle.
“There is a great deal of research going on nowadays into the problem of what to do with all the captured CO2,” Evans said. “It seems possible that we could eventually use solar energy to split hydrogen from water, and then combine that hydrogen with the CO2 to make more formic acid. Combined with ALF, that’s a solution that would help the planet.”
This simple material could scrub carbon dioxide from power plant smokestacks
The post This simple material could scrub carbon dioxide from power plant smokestacks first appeared on .
)
[1] => Array
(
[title] => How ancient fish colonized the deep sea
[link] => https://coolnspicy.com/science/how-ancient-fish-colonized-the-deep-sea/
[dc] => Array
(
[creator] => Michael Steiner
)
[pubdate] => Thu, 03 Nov 2022 08:15:56 +0000
[category] => scienceancientcolonizeddeepfishsea
[guid] => https://coolnspicy.com/?p=297
[description] => Journal Reference: Elizabeth Christina Miller, Christopher M. Martinez, Sarah T. Friedman, Peter C. Wainwright, Samantha A. Price, Luke Tornabene. Alternating regimes of shallow and deep-sea diversification explain a species-richness paradox in marine fishes. Proceedings of the National Academy of Sciences, 2022; 119 (43) DOI: 10.1073/pnas.2123544119 “It’s easy to look at shallow habitats like coral reefs, ... Read more
The post How ancient fish colonized the deep sea first appeared on .
[content] => Array
(
[encoded] =>
Journal Reference:
- Elizabeth Christina Miller, Christopher M. Martinez, Sarah T. Friedman, Peter C. Wainwright, Samantha A. Price, Luke Tornabene. Alternating regimes of shallow and deep-sea diversification explain a species-richness paradox in marine fishes. Proceedings of the National Academy of Sciences, 2022; 119 (43) DOI: 10.1073/pnas.2123544119
“It’s easy to look at shallow habitats like coral reefs, which are very diverse and exciting, and assume that they’ve always been that way,” said Miller, who completed the study as a postdoctoral researcher in the UW School of Aquatic and Fishery Sciences and is now a postdoctoral fellow at the University of Oklahoma. “These results really challenge that assumption, and help us understand how fish species have adapted to major changes to the climate.”
The deep sea is typically defined as anything below about 650 feet, the depth at which there is no longer enough sunlight for photosynthesis to occur. That means there is far less food and warmth than in the shallows, making it a difficult place to live. But by analyzing the relationships of fish using their genetic records going back 200 million years, Miller was able to identify a surprising evolutionary pattern: the speciation rates — that is, how quickly new species evolved — flip-flopped over time. There were periods lasting tens of millions of years when new species were evolving faster in the deep sea than in more shallow areas.
In some ways, this discovery raised more questions than it answered. What was causing fish to prefer one habitat over another? What made some fish able to move into the deep sea more easily than others? And how did these ancient shifts help create the diversity of species we have today?
When Miller mapped these flip-flopping speciation rates onto a timeline of Earth’s history, she was able to identify three major events that likely played a role.
“The first was the breakup of Pangea, which occurred between 200 and 150 million years ago,” said Miller. “That created new coastlines and new oceans, which meant there were more opportunities for fishes to move from shallow to deep water. There were suddenly a lot more access points.”
Next was the Cretaceous Hot Greenhouse period, which occurred approximately 100 million years ago and marked one of the warmest eras in Earth’s history. During this time, many continents were flooded due to sea-level rise, creating a large number of new, shallow areas across the earth.
“It was around this period that we really see shallow-water fishes take off and diversify,” said Miller. “We can trace a lot of the species diversity we see in the shallows today to this time.”
The third event was yet another major climatic change about 15 million years ago, known as the middle Miocene climatic transition. This was caused by further shifting of the continents, which caused major changes in ocean circulation and cooled the planet — all the way down to the deep sea.
“Around this time we see deep-sea speciation rates really speed up,” Miller said. “This was especially driven by cold-water fishes. A lot of the species you see today off the coasts of Washington and Alaska diversified during this time.”
But climate changes alone don’t explain how fish came to colonize the deep sea in the first place. Not every species has the right combination of traits to survive in deeper water and make use of the relatively limited resources beyond the reach of sunlight.
“To evolve into a new species in the deep sea, first you have to get there,” said Miller. “What we found was that not only were the speciation rates flip-flopping through time, but what the deep-sea fishes looked like was as well.”
The earliest fish that were able to transition into the deep sea tended to have large jaws. These likely gave them more opportunities to catch food, which can be scarce at depth. The researchers found that much later in history, fish that had longer, tapered tails tended to be most successful at making the transition to deep water. This allowed them to conserve energy by scooting along the seafloor instead of swimming in the water column.
“If you look at who lives in the deep sea today, some species have a tapered body and others have big, scary, toothy jaws,” Miller said. “Those two body plans represent ancestors that colonized the deep sea millions of years apart.”
While these events might seem like ancient history, they may be able to teach us about how today’s changing climate will affect life in our oceans. Miller hopes that future research can build on these findings and investigate how modern deep-sea fish will respond to climate change, and potentially inform conservation efforts.
“What we learned from this study is that deep-sea fishes tend to do well when oceans are colder, but with climate change, oceans are getting warmer,” she said. “We can expect that this is really going to impact fish in the deep-sea in the coming years.”
Co-authors are Luke Tornabene at the UW; Christopher Martinez at UC Irvine; Sarah Friedman at the NOAA Alaska Fisheries Science Center; Peter Wainwright at UC Davis; and Samantha Price at Clemson University.
This research was funded by the National Science Foundation.
Journal Reference:
- Elizabeth Christina Miller, Christopher M. Martinez, Sarah T. Friedman, Peter C. Wainwright, Samantha A. Price, Luke Tornabene. Alternating regimes of shallow and deep-sea diversification explain a species-richness paradox in marine fishes. Proceedings of the National Academy of Sciences, 2022; 119 (43) DOI: 10.1073/pnas.2123544119
“It’s easy to look at shallow habitats like coral reefs, which are very diverse and exciting, and assume that they’ve always been that way,” said Miller, who completed the study as a postdoctoral researcher in the UW School of Aquatic and Fishery Sciences and is now a postdoctoral fellow at the University of Oklahoma. “These results really challenge that assumption, and help us understand how fish species have adapted to major changes to the climate.”
The deep sea is typically defined as anything below about 650 feet, the depth at which there is no longer enough sunlight for photosynthesis to occur. That means there is far less food and warmth than in the shallows, making it a difficult place to live. But by analyzing the relationships of fish using their genetic records going back 200 million years, Miller was able to identify a surprising evolutionary pattern: the speciation rates — that is, how quickly new species evolved — flip-flopped over time. There were periods lasting tens of millions of years when new species were evolving faster in the deep sea than in more shallow areas.
In some ways, this discovery raised more questions than it answered. What was causing fish to prefer one habitat over another? What made some fish able to move into the deep sea more easily than others? And how did these ancient shifts help create the diversity of species we have today?
When Miller mapped these flip-flopping speciation rates onto a timeline of Earth’s history, she was able to identify three major events that likely played a role.
“The first was the breakup of Pangea, which occurred between 200 and 150 million years ago,” said Miller. “That created new coastlines and new oceans, which meant there were more opportunities for fishes to move from shallow to deep water. There were suddenly a lot more access points.”
Next was the Cretaceous Hot Greenhouse period, which occurred approximately 100 million years ago and marked one of the warmest eras in Earth’s history. During this time, many continents were flooded due to sea-level rise, creating a large number of new, shallow areas across the earth.
“It was around this period that we really see shallow-water fishes take off and diversify,” said Miller. “We can trace a lot of the species diversity we see in the shallows today to this time.”
The third event was yet another major climatic change about 15 million years ago, known as the middle Miocene climatic transition. This was caused by further shifting of the continents, which caused major changes in ocean circulation and cooled the planet — all the way down to the deep sea.
“Around this time we see deep-sea speciation rates really speed up,” Miller said. “This was especially driven by cold-water fishes. A lot of the species you see today off the coasts of Washington and Alaska diversified during this time.”
But climate changes alone don’t explain how fish came to colonize the deep sea in the first place. Not every species has the right combination of traits to survive in deeper water and make use of the relatively limited resources beyond the reach of sunlight.
“To evolve into a new species in the deep sea, first you have to get there,” said Miller. “What we found was that not only were the speciation rates flip-flopping through time, but what the deep-sea fishes looked like was as well.”
The earliest fish that were able to transition into the deep sea tended to have large jaws. These likely gave them more opportunities to catch food, which can be scarce at depth. The researchers found that much later in history, fish that had longer, tapered tails tended to be most successful at making the transition to deep water. This allowed them to conserve energy by scooting along the seafloor instead of swimming in the water column.
“If you look at who lives in the deep sea today, some species have a tapered body and others have big, scary, toothy jaws,” Miller said. “Those two body plans represent ancestors that colonized the deep sea millions of years apart.”
While these events might seem like ancient history, they may be able to teach us about how today’s changing climate will affect life in our oceans. Miller hopes that future research can build on these findings and investigate how modern deep-sea fish will respond to climate change, and potentially inform conservation efforts.
“What we learned from this study is that deep-sea fishes tend to do well when oceans are colder, but with climate change, oceans are getting warmer,” she said. “We can expect that this is really going to impact fish in the deep-sea in the coming years.”
Co-authors are Luke Tornabene at the UW; Christopher Martinez at UC Irvine; Sarah Friedman at the NOAA Alaska Fisheries Science Center; Peter Wainwright at UC Davis; and Samantha Price at Clemson University.
This research was funded by the National Science Foundation.
How ancient fish colonized the deep sea
The post How ancient fish colonized the deep sea first appeared on .
)
[summary] => Journal Reference: Elizabeth Christina Miller, Christopher M. Martinez, Sarah T. Friedman, Peter C. Wainwright, Samantha A. Price, Luke Tornabene. Alternating regimes of shallow and deep-sea diversification explain a species-richness paradox in marine fishes. Proceedings of the National Academy of Sciences, 2022; 119 (43) DOI: 10.1073/pnas.2123544119 “It’s easy to look at shallow habitats like coral reefs, ... Read more
The post How ancient fish colonized the deep sea first appeared on .
[atom_content] =>
Journal Reference:
- Elizabeth Christina Miller, Christopher M. Martinez, Sarah T. Friedman, Peter C. Wainwright, Samantha A. Price, Luke Tornabene. Alternating regimes of shallow and deep-sea diversification explain a species-richness paradox in marine fishes. Proceedings of the National Academy of Sciences, 2022; 119 (43) DOI: 10.1073/pnas.2123544119
“It’s easy to look at shallow habitats like coral reefs, which are very diverse and exciting, and assume that they’ve always been that way,” said Miller, who completed the study as a postdoctoral researcher in the UW School of Aquatic and Fishery Sciences and is now a postdoctoral fellow at the University of Oklahoma. “These results really challenge that assumption, and help us understand how fish species have adapted to major changes to the climate.”
The deep sea is typically defined as anything below about 650 feet, the depth at which there is no longer enough sunlight for photosynthesis to occur. That means there is far less food and warmth than in the shallows, making it a difficult place to live. But by analyzing the relationships of fish using their genetic records going back 200 million years, Miller was able to identify a surprising evolutionary pattern: the speciation rates — that is, how quickly new species evolved — flip-flopped over time. There were periods lasting tens of millions of years when new species were evolving faster in the deep sea than in more shallow areas.
In some ways, this discovery raised more questions than it answered. What was causing fish to prefer one habitat over another? What made some fish able to move into the deep sea more easily than others? And how did these ancient shifts help create the diversity of species we have today?
When Miller mapped these flip-flopping speciation rates onto a timeline of Earth’s history, she was able to identify three major events that likely played a role.
“The first was the breakup of Pangea, which occurred between 200 and 150 million years ago,” said Miller. “That created new coastlines and new oceans, which meant there were more opportunities for fishes to move from shallow to deep water. There were suddenly a lot more access points.”
Next was the Cretaceous Hot Greenhouse period, which occurred approximately 100 million years ago and marked one of the warmest eras in Earth’s history. During this time, many continents were flooded due to sea-level rise, creating a large number of new, shallow areas across the earth.
“It was around this period that we really see shallow-water fishes take off and diversify,” said Miller. “We can trace a lot of the species diversity we see in the shallows today to this time.”
The third event was yet another major climatic change about 15 million years ago, known as the middle Miocene climatic transition. This was caused by further shifting of the continents, which caused major changes in ocean circulation and cooled the planet — all the way down to the deep sea.
“Around this time we see deep-sea speciation rates really speed up,” Miller said. “This was especially driven by cold-water fishes. A lot of the species you see today off the coasts of Washington and Alaska diversified during this time.”
But climate changes alone don’t explain how fish came to colonize the deep sea in the first place. Not every species has the right combination of traits to survive in deeper water and make use of the relatively limited resources beyond the reach of sunlight.
“To evolve into a new species in the deep sea, first you have to get there,” said Miller. “What we found was that not only were the speciation rates flip-flopping through time, but what the deep-sea fishes looked like was as well.”
The earliest fish that were able to transition into the deep sea tended to have large jaws. These likely gave them more opportunities to catch food, which can be scarce at depth. The researchers found that much later in history, fish that had longer, tapered tails tended to be most successful at making the transition to deep water. This allowed them to conserve energy by scooting along the seafloor instead of swimming in the water column.
“If you look at who lives in the deep sea today, some species have a tapered body and others have big, scary, toothy jaws,” Miller said. “Those two body plans represent ancestors that colonized the deep sea millions of years apart.”
While these events might seem like ancient history, they may be able to teach us about how today’s changing climate will affect life in our oceans. Miller hopes that future research can build on these findings and investigate how modern deep-sea fish will respond to climate change, and potentially inform conservation efforts.
“What we learned from this study is that deep-sea fishes tend to do well when oceans are colder, but with climate change, oceans are getting warmer,” she said. “We can expect that this is really going to impact fish in the deep-sea in the coming years.”
Co-authors are Luke Tornabene at the UW; Christopher Martinez at UC Irvine; Sarah Friedman at the NOAA Alaska Fisheries Science Center; Peter Wainwright at UC Davis; and Samantha Price at Clemson University.
This research was funded by the National Science Foundation.
Journal Reference:
- Elizabeth Christina Miller, Christopher M. Martinez, Sarah T. Friedman, Peter C. Wainwright, Samantha A. Price, Luke Tornabene. Alternating regimes of shallow and deep-sea diversification explain a species-richness paradox in marine fishes. Proceedings of the National Academy of Sciences, 2022; 119 (43) DOI: 10.1073/pnas.2123544119
“It’s easy to look at shallow habitats like coral reefs, which are very diverse and exciting, and assume that they’ve always been that way,” said Miller, who completed the study as a postdoctoral researcher in the UW School of Aquatic and Fishery Sciences and is now a postdoctoral fellow at the University of Oklahoma. “These results really challenge that assumption, and help us understand how fish species have adapted to major changes to the climate.”
The deep sea is typically defined as anything below about 650 feet, the depth at which there is no longer enough sunlight for photosynthesis to occur. That means there is far less food and warmth than in the shallows, making it a difficult place to live. But by analyzing the relationships of fish using their genetic records going back 200 million years, Miller was able to identify a surprising evolutionary pattern: the speciation rates — that is, how quickly new species evolved — flip-flopped over time. There were periods lasting tens of millions of years when new species were evolving faster in the deep sea than in more shallow areas.
In some ways, this discovery raised more questions than it answered. What was causing fish to prefer one habitat over another? What made some fish able to move into the deep sea more easily than others? And how did these ancient shifts help create the diversity of species we have today?
When Miller mapped these flip-flopping speciation rates onto a timeline of Earth’s history, she was able to identify three major events that likely played a role.
“The first was the breakup of Pangea, which occurred between 200 and 150 million years ago,” said Miller. “That created new coastlines and new oceans, which meant there were more opportunities for fishes to move from shallow to deep water. There were suddenly a lot more access points.”
Next was the Cretaceous Hot Greenhouse period, which occurred approximately 100 million years ago and marked one of the warmest eras in Earth’s history. During this time, many continents were flooded due to sea-level rise, creating a large number of new, shallow areas across the earth.
“It was around this period that we really see shallow-water fishes take off and diversify,” said Miller. “We can trace a lot of the species diversity we see in the shallows today to this time.”
The third event was yet another major climatic change about 15 million years ago, known as the middle Miocene climatic transition. This was caused by further shifting of the continents, which caused major changes in ocean circulation and cooled the planet — all the way down to the deep sea.
“Around this time we see deep-sea speciation rates really speed up,” Miller said. “This was especially driven by cold-water fishes. A lot of the species you see today off the coasts of Washington and Alaska diversified during this time.”
But climate changes alone don’t explain how fish came to colonize the deep sea in the first place. Not every species has the right combination of traits to survive in deeper water and make use of the relatively limited resources beyond the reach of sunlight.
“To evolve into a new species in the deep sea, first you have to get there,” said Miller. “What we found was that not only were the speciation rates flip-flopping through time, but what the deep-sea fishes looked like was as well.”
The earliest fish that were able to transition into the deep sea tended to have large jaws. These likely gave them more opportunities to catch food, which can be scarce at depth. The researchers found that much later in history, fish that had longer, tapered tails tended to be most successful at making the transition to deep water. This allowed them to conserve energy by scooting along the seafloor instead of swimming in the water column.
“If you look at who lives in the deep sea today, some species have a tapered body and others have big, scary, toothy jaws,” Miller said. “Those two body plans represent ancestors that colonized the deep sea millions of years apart.”
While these events might seem like ancient history, they may be able to teach us about how today’s changing climate will affect life in our oceans. Miller hopes that future research can build on these findings and investigate how modern deep-sea fish will respond to climate change, and potentially inform conservation efforts.
“What we learned from this study is that deep-sea fishes tend to do well when oceans are colder, but with climate change, oceans are getting warmer,” she said. “We can expect that this is really going to impact fish in the deep-sea in the coming years.”
Co-authors are Luke Tornabene at the UW; Christopher Martinez at UC Irvine; Sarah Friedman at the NOAA Alaska Fisheries Science Center; Peter Wainwright at UC Davis; and Samantha Price at Clemson University.
This research was funded by the National Science Foundation.
How ancient fish colonized the deep sea
The post How ancient fish colonized the deep sea first appeared on .
)
[2] => Array
(
[title] => Ambrosia beetles breed and maintain their own food fungi
[link] => https://coolnspicy.com/science/ambrosia-beetles-breed-and-maintain-their-own-food-fungi/
[dc] => Array
(
[creator] => Michael Steiner
)
[pubdate] => Thu, 03 Nov 2022 08:11:06 +0000
[category] => scienceAmbrosiabeetlesbreedfoodfungimaintain
[guid] => https://coolnspicy.com/?p=295
[description] => Journal Reference: Janina M. C. Diehl, Vienna Kowallik, Alexander Keller, Peter H. W. Biedermann. First experimental evidence for active farming in ambrosia beetles and strong heredity of garden microbiomes. Proceedings of the Royal Society B: Biological Sciences, 2022; 289 (1986) DOI: 10.1098/rspb.2022.1458 Fungal coatings in wooden tunnels Ambrosia beetles feed on special fungal coatings that ... Read more
The post Ambrosia beetles breed and maintain their own food fungi first appeared on .
[content] => Array
(
[encoded] =>
Journal Reference:
- Janina M. C. Diehl, Vienna Kowallik, Alexander Keller, Peter H. W. Biedermann. First experimental evidence for active farming in ambrosia beetles and strong heredity of garden microbiomes. Proceedings of the Royal Society B: Biological Sciences, 2022; 289 (1986) DOI: 10.1098/rspb.2022.1458
Fungal coatings in wooden tunnels
Ambrosia beetles feed on special fungal coatings that grow in the tunnels they bore into old wood. To early naturalists, these coverings seemed like divine ambrosia, which is how the beetles got their name. Due to their social and hygienic behavior, it has long been assumed that they actively care for their fungi, but so far, such agricultural abilities have only been demonstrated in some termites and leafcutter ants.
Genetic analysis of fungus gardens
Diehl has now also succeeded in doing this for ambrosia beetles: In the laboratory, she had mother beetles of the little wood borer establish nests with offspring, in which the typical fungal gardens formed. She then removed the nurturing individuals from some of the nests and left them in others. Genetic analysis of bacterial and fungal communities of the fungal gardens after 40 days showed that the presence of the beetles had greatly altered the fungal community.
“You might have expected there to be fewer food fungi in the nests with beetles because they were being eaten, but in fact, the opposite was true; here the fungal composition was clearly shifted toward food fungi,” says Diehl. In the nests without nurturing beetles, on the other hand, the proportion of weed fungi was significantly higher. The composition of the bacteria also differed.
Beetles probably use antibiotic-forming bacteria
“These results support the existence of active farming in ambrosia beetles, although the exact mechanisms controlling the fungal community need further investigation,” adds Biedermann. He says there is evidence that the beetles use specific bacteria that produce antibiotic substances. These, in turn, could inhibit the growth of the weed fungi.
Social behavior probably also plays an important role; the entire group of beetles in the nest, including the larvae, work together to care for the fungi. This creates a close symbiosis between beetles and fungi: “Each ambrosia beetle species has its own food fungus. Neither can survive without the other.”
60 million years of experience
Economically relevant bark beetles, such as the spruce bark beetle (Ips typographus), also have similar symbioses with fungi, and understanding them could help control the beetles better in the future. Further research into how exactly ambrosia beetles suppress the growth of weed fungi could also provide worthwhile insights for human agriculture, which is struggling with resistance, for example, says Biedermann. “It’s highly exciting for us to see how nature has been doing this for 60 million years. Presumably, we humans can still learn something from these mechanisms.”
Journal Reference:
- Janina M. C. Diehl, Vienna Kowallik, Alexander Keller, Peter H. W. Biedermann. First experimental evidence for active farming in ambrosia beetles and strong heredity of garden microbiomes. Proceedings of the Royal Society B: Biological Sciences, 2022; 289 (1986) DOI: 10.1098/rspb.2022.1458
Fungal coatings in wooden tunnels
Ambrosia beetles feed on special fungal coatings that grow in the tunnels they bore into old wood. To early naturalists, these coverings seemed like divine ambrosia, which is how the beetles got their name. Due to their social and hygienic behavior, it has long been assumed that they actively care for their fungi, but so far, such agricultural abilities have only been demonstrated in some termites and leafcutter ants.
Genetic analysis of fungus gardens
Diehl has now also succeeded in doing this for ambrosia beetles: In the laboratory, she had mother beetles of the little wood borer establish nests with offspring, in which the typical fungal gardens formed. She then removed the nurturing individuals from some of the nests and left them in others. Genetic analysis of bacterial and fungal communities of the fungal gardens after 40 days showed that the presence of the beetles had greatly altered the fungal community.
“You might have expected there to be fewer food fungi in the nests with beetles because they were being eaten, but in fact, the opposite was true; here the fungal composition was clearly shifted toward food fungi,” says Diehl. In the nests without nurturing beetles, on the other hand, the proportion of weed fungi was significantly higher. The composition of the bacteria also differed.
Beetles probably use antibiotic-forming bacteria
“These results support the existence of active farming in ambrosia beetles, although the exact mechanisms controlling the fungal community need further investigation,” adds Biedermann. He says there is evidence that the beetles use specific bacteria that produce antibiotic substances. These, in turn, could inhibit the growth of the weed fungi.
Social behavior probably also plays an important role; the entire group of beetles in the nest, including the larvae, work together to care for the fungi. This creates a close symbiosis between beetles and fungi: “Each ambrosia beetle species has its own food fungus. Neither can survive without the other.”
60 million years of experience
Economically relevant bark beetles, such as the spruce bark beetle (Ips typographus), also have similar symbioses with fungi, and understanding them could help control the beetles better in the future. Further research into how exactly ambrosia beetles suppress the growth of weed fungi could also provide worthwhile insights for human agriculture, which is struggling with resistance, for example, says Biedermann. “It’s highly exciting for us to see how nature has been doing this for 60 million years. Presumably, we humans can still learn something from these mechanisms.”
original title
The post Ambrosia beetles breed and maintain their own food fungi first appeared on .
)
[summary] => Journal Reference: Janina M. C. Diehl, Vienna Kowallik, Alexander Keller, Peter H. W. Biedermann. First experimental evidence for active farming in ambrosia beetles and strong heredity of garden microbiomes. Proceedings of the Royal Society B: Biological Sciences, 2022; 289 (1986) DOI: 10.1098/rspb.2022.1458 Fungal coatings in wooden tunnels Ambrosia beetles feed on special fungal coatings that ... Read more
The post Ambrosia beetles breed and maintain their own food fungi first appeared on .
[atom_content] =>
Journal Reference:
- Janina M. C. Diehl, Vienna Kowallik, Alexander Keller, Peter H. W. Biedermann. First experimental evidence for active farming in ambrosia beetles and strong heredity of garden microbiomes. Proceedings of the Royal Society B: Biological Sciences, 2022; 289 (1986) DOI: 10.1098/rspb.2022.1458
Fungal coatings in wooden tunnels
Ambrosia beetles feed on special fungal coatings that grow in the tunnels they bore into old wood. To early naturalists, these coverings seemed like divine ambrosia, which is how the beetles got their name. Due to their social and hygienic behavior, it has long been assumed that they actively care for their fungi, but so far, such agricultural abilities have only been demonstrated in some termites and leafcutter ants.
Genetic analysis of fungus gardens
Diehl has now also succeeded in doing this for ambrosia beetles: In the laboratory, she had mother beetles of the little wood borer establish nests with offspring, in which the typical fungal gardens formed. She then removed the nurturing individuals from some of the nests and left them in others. Genetic analysis of bacterial and fungal communities of the fungal gardens after 40 days showed that the presence of the beetles had greatly altered the fungal community.
“You might have expected there to be fewer food fungi in the nests with beetles because they were being eaten, but in fact, the opposite was true; here the fungal composition was clearly shifted toward food fungi,” says Diehl. In the nests without nurturing beetles, on the other hand, the proportion of weed fungi was significantly higher. The composition of the bacteria also differed.
Beetles probably use antibiotic-forming bacteria
“These results support the existence of active farming in ambrosia beetles, although the exact mechanisms controlling the fungal community need further investigation,” adds Biedermann. He says there is evidence that the beetles use specific bacteria that produce antibiotic substances. These, in turn, could inhibit the growth of the weed fungi.
Social behavior probably also plays an important role; the entire group of beetles in the nest, including the larvae, work together to care for the fungi. This creates a close symbiosis between beetles and fungi: “Each ambrosia beetle species has its own food fungus. Neither can survive without the other.”
60 million years of experience
Economically relevant bark beetles, such as the spruce bark beetle (Ips typographus), also have similar symbioses with fungi, and understanding them could help control the beetles better in the future. Further research into how exactly ambrosia beetles suppress the growth of weed fungi could also provide worthwhile insights for human agriculture, which is struggling with resistance, for example, says Biedermann. “It’s highly exciting for us to see how nature has been doing this for 60 million years. Presumably, we humans can still learn something from these mechanisms.”
Journal Reference:
- Janina M. C. Diehl, Vienna Kowallik, Alexander Keller, Peter H. W. Biedermann. First experimental evidence for active farming in ambrosia beetles and strong heredity of garden microbiomes. Proceedings of the Royal Society B: Biological Sciences, 2022; 289 (1986) DOI: 10.1098/rspb.2022.1458
Fungal coatings in wooden tunnels
Ambrosia beetles feed on special fungal coatings that grow in the tunnels they bore into old wood. To early naturalists, these coverings seemed like divine ambrosia, which is how the beetles got their name. Due to their social and hygienic behavior, it has long been assumed that they actively care for their fungi, but so far, such agricultural abilities have only been demonstrated in some termites and leafcutter ants.
Genetic analysis of fungus gardens
Diehl has now also succeeded in doing this for ambrosia beetles: In the laboratory, she had mother beetles of the little wood borer establish nests with offspring, in which the typical fungal gardens formed. She then removed the nurturing individuals from some of the nests and left them in others. Genetic analysis of bacterial and fungal communities of the fungal gardens after 40 days showed that the presence of the beetles had greatly altered the fungal community.
“You might have expected there to be fewer food fungi in the nests with beetles because they were being eaten, but in fact, the opposite was true; here the fungal composition was clearly shifted toward food fungi,” says Diehl. In the nests without nurturing beetles, on the other hand, the proportion of weed fungi was significantly higher. The composition of the bacteria also differed.
Beetles probably use antibiotic-forming bacteria
“These results support the existence of active farming in ambrosia beetles, although the exact mechanisms controlling the fungal community need further investigation,” adds Biedermann. He says there is evidence that the beetles use specific bacteria that produce antibiotic substances. These, in turn, could inhibit the growth of the weed fungi.
Social behavior probably also plays an important role; the entire group of beetles in the nest, including the larvae, work together to care for the fungi. This creates a close symbiosis between beetles and fungi: “Each ambrosia beetle species has its own food fungus. Neither can survive without the other.”
60 million years of experience
Economically relevant bark beetles, such as the spruce bark beetle (Ips typographus), also have similar symbioses with fungi, and understanding them could help control the beetles better in the future. Further research into how exactly ambrosia beetles suppress the growth of weed fungi could also provide worthwhile insights for human agriculture, which is struggling with resistance, for example, says Biedermann. “It’s highly exciting for us to see how nature has been doing this for 60 million years. Presumably, we humans can still learn something from these mechanisms.”
original title
The post Ambrosia beetles breed and maintain their own food fungi first appeared on .
)
[3] => Array
(
[title] => Glowing fossils: Fluorescence reveals color patterns of earliest scallops
[link] => https://coolnspicy.com/science/glowing-fossils-fluorescence-reveals-color-patterns-of-earliest-scallops/
[dc] => Array
(
[creator] => Michael Steiner
)
[pubdate] => Thu, 03 Nov 2022 08:09:49 +0000
[category] => sciencecolorearliestFluorescencefossilsGlowingpatternsrevealsscallops
[guid] => https://coolnspicy.com/?p=293
[description] => Journal Reference: Klaus Wolkenstein. Fluorescent colour patterns in the basal pectinid Pleuronectites from the Middle Triassic of Central Europe: origin, fate and taxonomic implications of fluorescence. Palaeontology, 2022; 65 (5) DOI: 10.1111/pala.12625 In fossils from the Mesozoic Era, traces of colour patterns are very rarely observed. However, the investigation with UV light of scallops from ... Read more
The post Glowing fossils: Fluorescence reveals color patterns of earliest scallops first appeared on .
[content] => Array
(
[encoded] =>
Journal Reference:
- Klaus Wolkenstein. Fluorescent colour patterns in the basal pectinid Pleuronectites from the Middle Triassic of Central Europe: origin, fate and taxonomic implications of fluorescence. Palaeontology, 2022; 65 (5) DOI: 10.1111/pala.12625
In fossils from the Mesozoic Era, traces of colour patterns are very rarely observed. However, the investigation with UV light of scallops from the Triassic period — right from the beginning of the Mesozoic Era — shows that colour patterns are preserved much more frequently than previously thought. UV light, which is invisible to the human eye, excites organic compounds in the fossils causing them to glow. This reveals a surprising variety of colour patterns: different variations of stripes, zigzags and flame patterns. The diversity of colour patterns is similar to those of today’s seashells found on a beach.
However, the colour patterns of today’s scallops do not show any fluorescence. “In the case of the Triassic shells, fluorescent compounds were only formed in the course of fossilisation through oxidation of the original pigments,” explains Dr Klaus Wolkenstein from the Geosciences Centre at the University of Göttingen, who is currently carrying out research at the University of Bonn. Surprisingly, the fossil shells show different fluorescent colours, depending on the region where they were found. “The colour spectrum ranges from yellow to red with all the transitions in between, which suggests that there were clear regional differences in the fossilisation of these scallops,” adds Wolkenstein.
Journal Reference:
- Klaus Wolkenstein. Fluorescent colour patterns in the basal pectinid Pleuronectites from the Middle Triassic of Central Europe: origin, fate and taxonomic implications of fluorescence. Palaeontology, 2022; 65 (5) DOI: 10.1111/pala.12625
In fossils from the Mesozoic Era, traces of colour patterns are very rarely observed. However, the investigation with UV light of scallops from the Triassic period — right from the beginning of the Mesozoic Era — shows that colour patterns are preserved much more frequently than previously thought. UV light, which is invisible to the human eye, excites organic compounds in the fossils causing them to glow. This reveals a surprising variety of colour patterns: different variations of stripes, zigzags and flame patterns. The diversity of colour patterns is similar to those of today’s seashells found on a beach.
However, the colour patterns of today’s scallops do not show any fluorescence. “In the case of the Triassic shells, fluorescent compounds were only formed in the course of fossilisation through oxidation of the original pigments,” explains Dr Klaus Wolkenstein from the Geosciences Centre at the University of Göttingen, who is currently carrying out research at the University of Bonn. Surprisingly, the fossil shells show different fluorescent colours, depending on the region where they were found. “The colour spectrum ranges from yellow to red with all the transitions in between, which suggests that there were clear regional differences in the fossilisation of these scallops,” adds Wolkenstein.
Source link
The post Glowing fossils: Fluorescence reveals color patterns of earliest scallops first appeared on .
)
[summary] => Journal Reference: Klaus Wolkenstein. Fluorescent colour patterns in the basal pectinid Pleuronectites from the Middle Triassic of Central Europe: origin, fate and taxonomic implications of fluorescence. Palaeontology, 2022; 65 (5) DOI: 10.1111/pala.12625 In fossils from the Mesozoic Era, traces of colour patterns are very rarely observed. However, the investigation with UV light of scallops from ... Read more
The post Glowing fossils: Fluorescence reveals color patterns of earliest scallops first appeared on .
[atom_content] =>
Journal Reference:
- Klaus Wolkenstein. Fluorescent colour patterns in the basal pectinid Pleuronectites from the Middle Triassic of Central Europe: origin, fate and taxonomic implications of fluorescence. Palaeontology, 2022; 65 (5) DOI: 10.1111/pala.12625
In fossils from the Mesozoic Era, traces of colour patterns are very rarely observed. However, the investigation with UV light of scallops from the Triassic period — right from the beginning of the Mesozoic Era — shows that colour patterns are preserved much more frequently than previously thought. UV light, which is invisible to the human eye, excites organic compounds in the fossils causing them to glow. This reveals a surprising variety of colour patterns: different variations of stripes, zigzags and flame patterns. The diversity of colour patterns is similar to those of today’s seashells found on a beach.
However, the colour patterns of today’s scallops do not show any fluorescence. “In the case of the Triassic shells, fluorescent compounds were only formed in the course of fossilisation through oxidation of the original pigments,” explains Dr Klaus Wolkenstein from the Geosciences Centre at the University of Göttingen, who is currently carrying out research at the University of Bonn. Surprisingly, the fossil shells show different fluorescent colours, depending on the region where they were found. “The colour spectrum ranges from yellow to red with all the transitions in between, which suggests that there were clear regional differences in the fossilisation of these scallops,” adds Wolkenstein.
Journal Reference:
- Klaus Wolkenstein. Fluorescent colour patterns in the basal pectinid Pleuronectites from the Middle Triassic of Central Europe: origin, fate and taxonomic implications of fluorescence. Palaeontology, 2022; 65 (5) DOI: 10.1111/pala.12625
In fossils from the Mesozoic Era, traces of colour patterns are very rarely observed. However, the investigation with UV light of scallops from the Triassic period — right from the beginning of the Mesozoic Era — shows that colour patterns are preserved much more frequently than previously thought. UV light, which is invisible to the human eye, excites organic compounds in the fossils causing them to glow. This reveals a surprising variety of colour patterns: different variations of stripes, zigzags and flame patterns. The diversity of colour patterns is similar to those of today’s seashells found on a beach.
However, the colour patterns of today’s scallops do not show any fluorescence. “In the case of the Triassic shells, fluorescent compounds were only formed in the course of fossilisation through oxidation of the original pigments,” explains Dr Klaus Wolkenstein from the Geosciences Centre at the University of Göttingen, who is currently carrying out research at the University of Bonn. Surprisingly, the fossil shells show different fluorescent colours, depending on the region where they were found. “The colour spectrum ranges from yellow to red with all the transitions in between, which suggests that there were clear regional differences in the fossilisation of these scallops,” adds Wolkenstein.
Source link
The post Glowing fossils: Fluorescence reveals color patterns of earliest scallops first appeared on .
)
[4] => Array
(
[title] => Unveiling the dimensionality of complex networks through hyperbolic geometry
[link] => https://coolnspicy.com/science/unveiling-the-dimensionality-of-complex-networks-through-hyperbolic-geometry/
[dc] => Array
(
[creator] => Michael Steiner
)
[pubdate] => Thu, 03 Nov 2022 08:09:34 +0000
[category] => sciencecomplexdimensionalitygeometryhyperbolicnetworksUnveiling
[guid] => https://coolnspicy.com/?p=291
[description] => Journal Reference: Pedro Almagro, Marián Boguñá, M. Ángeles Serrano. Detecting the ultra low dimensionality of real networks. Nature Communications, 2022; 13 (1) DOI: 10.1038/s41467-022-33685-z Among the authors of the study are the researchers M. Ángeles Serrano and Marián Boguñá, from the Faculty of Physics and the Institute of Complex Systems of the UB (UBICS), and ... Read more
The post Unveiling the dimensionality of complex networks through hyperbolic geometry first appeared on .
[content] => Array
(
[encoded] =>
Journal Reference:
- Pedro Almagro, Marián Boguñá, M. Ángeles Serrano. Detecting the ultra low dimensionality of real networks. Nature Communications, 2022; 13 (1) DOI: 10.1038/s41467-022-33685-z
Among the authors of the study are the researchers M. Ángeles Serrano and Marián Boguñá, from the Faculty of Physics and the Institute of Complex Systems of the UB (UBICS), and Pedro Almargo, from the Higher Technical School of Engineering of the University of Sevilla. The research study provides a multidimensional hyperbolic model of complex networks that reproduces its connectivity, with an ultra-low and customizable dimensionality for each specific network. This enables a better characterization of its structure — e.g. at a community scale — and the improvement of its predictive capability.
The study reveals unexpected regularities, such as the extremely low dimensions of molecular networks associated with biological tissues; the slightly higher dimensionality required by social networks and the Internet; and the discovery that brain connectomes are close to three dimensions in their automatic organisation.
Hyperbolic versus Euclidean geometry
The intrinsic geometry of data sets or complex networks is not obvious, which becomes an obstacle in determining the dimensionality of real networks. Another challenge is that the definition of distance has to be established according to their relational and connectivity structure, and this also requires sophisticated models.
Now, the new approach is based on the geometry of complex networks, and more specifically, on the configurational geometric model or SD model. “This model, which we have developed in previous work, describes the structure of complex networks based on fundamental principles,” says the lecturer M. Ángeles, ICREA researcher at the Department of Condensed Matter Physics of the UB.
“More specifically — he continues — , the model postulates a law of interconnection of the network elements (or nodes) that is gravitational, so nodes that are closer in a similarity space — of spherical geometry in D dimensions — and with more popularity — an extra dimension corresponding to the importance of the node — are more likely to establish connections.”
In the study, the similarity and popularity variables are combined to give rise to the hyperbolic geometry of the model, which emerges as the natural geometry representing the hierarchical architecture of complex networks.
In previous studies, the team had applied the simplest version of the one-dimensional SD model — the S1 model — to explain many typical features of real-world networks: the small-world property (the six degrees of separation), the heterogeneous distributions of the number of neighbours per node, and the high levels of transitive relationships (triangle connections that can be illustrated with the expression my friend’s friend is also my friend).
“In addition, the application of statistical inference techniques allows us to obtain real network maps in the hyperbolic plan that are congruent with the established model,” she says. “Beyond visualisation, these representations have been used in a multitude of tasks, including efficient navigation methods, the detection of self-similarity patterns, the detection of strongly interacting communities of nodes, and the implementation of a network renormalisation procedure that reveals hidden symmetries in the multi-scale organisation of complex networks and allows the production of network replicas at reduced or enlarged scales.”
Now, the team infers the dimensionality of the hyperbolic space underlying the real networks from properties that relate to the dimension of their geometry. In particular, the work measures the statistics of higher-order cycles (triangles, squares, pentagons) associated with the connections.
A methodology applicable to all complex networks
In computer science, the applied techniques are based on data that typically make definitions of similarity distance between their elements, an approach that involves the construction of graphs that are mapped onto a latent space of Euclidean features.
“Our estimates of the dimensionality of complex networks are well below our estimates based on Euclidean space, since hyperbolic space is better suited to represent the hierarchical structure of real complex networks. For example, the Internet only requires D = 7 dimensions to be mapped into the hyperbolic space of our model, whereas this name is multiplied by six and scales to D = 47 in one of the most recent techniques using Euclidean space,” says Professor Marián Boguñá.
In addition, techniques for mapping complex data usually assume a latent space, with a predetermined name of dimensions, or implement heuristic techniques to find a suitable value. Thus, the new method is based on a model that does not need the spatial mapping of the network to determine the dimension of its geometry.
In the field of network science, many methodologies use the shortest distances to study the connectivity structure of the network (shortest paths) as a metric space. However, these distances are strongly affected by the small-world property and do not provide a wide range of distance values.
“Our model uses a completely different definition of distance based on an underlying hyperbolic space, and we do not need to map the network. Our methodology is applicable to any real network or data series with complex structure and with a size that is typically thousands or tens of thousands of nodes but can reach hundreds of thousands in a reasonable computational time,” says M. Ángeles Serrano.
What is the real dimensionality of social networks and the Internet?
ocial networks and the Internet is higher (between 6 and 9) compared to networks in other domains, according to the study’s findings. However, it is still very low — 6 to 7 times lower — compared to that obtained by other methods. This reflects the fact that interactions in these systems are more complex and determined by a greater variety of factors.
On the other hand, friendship-based social networks are at the top of the dimensionality ranking. “This is an unexpected result, since one might think that friendship is a freer type of affective relationship, but our results link to the fact that homophily in human interactions is determined by a multitude of sociological factors such as age, gender, social class, beliefs, attitudes or interests,” says M. Ángeles Serrano.
In the case of the Internet, even though it is a technological network, its greater dimensionality reflects the fact that for an autonomous system, connecting does not mean only accessing the system, as one might think at first. On the contrary, many different factors influence the formation of these connections, and as a consequence, a variety of other relationships may be present (e.g., supplier-client, peer-to-peer, exchange-based peering, etc.).
“What is really surprising, both for social networks and the internet, is that our theoretical framework — which does not use any annotations about connections beyond their existence — is able to capture this multidimensional reality that is not explicit in our data,” concludes the team, which is currently working on constructing hyperbolic multidimensional maps of complex networks that are congruent with the theoretical framework established by the SD model.
Journal Reference:
- Pedro Almagro, Marián Boguñá, M. Ángeles Serrano. Detecting the ultra low dimensionality of real networks. Nature Communications, 2022; 13 (1) DOI: 10.1038/s41467-022-33685-z
Among the authors of the study are the researchers M. Ángeles Serrano and Marián Boguñá, from the Faculty of Physics and the Institute of Complex Systems of the UB (UBICS), and Pedro Almargo, from the Higher Technical School of Engineering of the University of Sevilla. The research study provides a multidimensional hyperbolic model of complex networks that reproduces its connectivity, with an ultra-low and customizable dimensionality for each specific network. This enables a better characterization of its structure — e.g. at a community scale — and the improvement of its predictive capability.
The study reveals unexpected regularities, such as the extremely low dimensions of molecular networks associated with biological tissues; the slightly higher dimensionality required by social networks and the Internet; and the discovery that brain connectomes are close to three dimensions in their automatic organisation.
Hyperbolic versus Euclidean geometry
The intrinsic geometry of data sets or complex networks is not obvious, which becomes an obstacle in determining the dimensionality of real networks. Another challenge is that the definition of distance has to be established according to their relational and connectivity structure, and this also requires sophisticated models.
Now, the new approach is based on the geometry of complex networks, and more specifically, on the configurational geometric model or SD model. “This model, which we have developed in previous work, describes the structure of complex networks based on fundamental principles,” says the lecturer M. Ángeles, ICREA researcher at the Department of Condensed Matter Physics of the UB.
“More specifically — he continues — , the model postulates a law of interconnection of the network elements (or nodes) that is gravitational, so nodes that are closer in a similarity space — of spherical geometry in D dimensions — and with more popularity — an extra dimension corresponding to the importance of the node — are more likely to establish connections.”
In the study, the similarity and popularity variables are combined to give rise to the hyperbolic geometry of the model, which emerges as the natural geometry representing the hierarchical architecture of complex networks.
In previous studies, the team had applied the simplest version of the one-dimensional SD model — the S1 model — to explain many typical features of real-world networks: the small-world property (the six degrees of separation), the heterogeneous distributions of the number of neighbours per node, and the high levels of transitive relationships (triangle connections that can be illustrated with the expression my friend’s friend is also my friend).
“In addition, the application of statistical inference techniques allows us to obtain real network maps in the hyperbolic plan that are congruent with the established model,” she says. “Beyond visualisation, these representations have been used in a multitude of tasks, including efficient navigation methods, the detection of self-similarity patterns, the detection of strongly interacting communities of nodes, and the implementation of a network renormalisation procedure that reveals hidden symmetries in the multi-scale organisation of complex networks and allows the production of network replicas at reduced or enlarged scales.”
Now, the team infers the dimensionality of the hyperbolic space underlying the real networks from properties that relate to the dimension of their geometry. In particular, the work measures the statistics of higher-order cycles (triangles, squares, pentagons) associated with the connections.
A methodology applicable to all complex networks
In computer science, the applied techniques are based on data that typically make definitions of similarity distance between their elements, an approach that involves the construction of graphs that are mapped onto a latent space of Euclidean features.
“Our estimates of the dimensionality of complex networks are well below our estimates based on Euclidean space, since hyperbolic space is better suited to represent the hierarchical structure of real complex networks. For example, the Internet only requires D = 7 dimensions to be mapped into the hyperbolic space of our model, whereas this name is multiplied by six and scales to D = 47 in one of the most recent techniques using Euclidean space,” says Professor Marián Boguñá.
In addition, techniques for mapping complex data usually assume a latent space, with a predetermined name of dimensions, or implement heuristic techniques to find a suitable value. Thus, the new method is based on a model that does not need the spatial mapping of the network to determine the dimension of its geometry.
In the field of network science, many methodologies use the shortest distances to study the connectivity structure of the network (shortest paths) as a metric space. However, these distances are strongly affected by the small-world property and do not provide a wide range of distance values.
“Our model uses a completely different definition of distance based on an underlying hyperbolic space, and we do not need to map the network. Our methodology is applicable to any real network or data series with complex structure and with a size that is typically thousands or tens of thousands of nodes but can reach hundreds of thousands in a reasonable computational time,” says M. Ángeles Serrano.
What is the real dimensionality of social networks and the Internet?
ocial networks and the Internet is higher (between 6 and 9) compared to networks in other domains, according to the study’s findings. However, it is still very low — 6 to 7 times lower — compared to that obtained by other methods. This reflects the fact that interactions in these systems are more complex and determined by a greater variety of factors.
On the other hand, friendship-based social networks are at the top of the dimensionality ranking. “This is an unexpected result, since one might think that friendship is a freer type of affective relationship, but our results link to the fact that homophily in human interactions is determined by a multitude of sociological factors such as age, gender, social class, beliefs, attitudes or interests,” says M. Ángeles Serrano.
In the case of the Internet, even though it is a technological network, its greater dimensionality reflects the fact that for an autonomous system, connecting does not mean only accessing the system, as one might think at first. On the contrary, many different factors influence the formation of these connections, and as a consequence, a variety of other relationships may be present (e.g., supplier-client, peer-to-peer, exchange-based peering, etc.).
“What is really surprising, both for social networks and the internet, is that our theoretical framework — which does not use any annotations about connections beyond their existence — is able to capture this multidimensional reality that is not explicit in our data,” concludes the team, which is currently working on constructing hyperbolic multidimensional maps of complex networks that are congruent with the theoretical framework established by the SD model.
Source link
The post Unveiling the dimensionality of complex networks through hyperbolic geometry first appeared on .
)
[summary] => Journal Reference: Pedro Almagro, Marián Boguñá, M. Ángeles Serrano. Detecting the ultra low dimensionality of real networks. Nature Communications, 2022; 13 (1) DOI: 10.1038/s41467-022-33685-z Among the authors of the study are the researchers M. Ángeles Serrano and Marián Boguñá, from the Faculty of Physics and the Institute of Complex Systems of the UB (UBICS), and ... Read more
The post Unveiling the dimensionality of complex networks through hyperbolic geometry first appeared on .
[atom_content] =>
Journal Reference:
- Pedro Almagro, Marián Boguñá, M. Ángeles Serrano. Detecting the ultra low dimensionality of real networks. Nature Communications, 2022; 13 (1) DOI: 10.1038/s41467-022-33685-z
Among the authors of the study are the researchers M. Ángeles Serrano and Marián Boguñá, from the Faculty of Physics and the Institute of Complex Systems of the UB (UBICS), and Pedro Almargo, from the Higher Technical School of Engineering of the University of Sevilla. The research study provides a multidimensional hyperbolic model of complex networks that reproduces its connectivity, with an ultra-low and customizable dimensionality for each specific network. This enables a better characterization of its structure — e.g. at a community scale — and the improvement of its predictive capability.
The study reveals unexpected regularities, such as the extremely low dimensions of molecular networks associated with biological tissues; the slightly higher dimensionality required by social networks and the Internet; and the discovery that brain connectomes are close to three dimensions in their automatic organisation.
Hyperbolic versus Euclidean geometry
The intrinsic geometry of data sets or complex networks is not obvious, which becomes an obstacle in determining the dimensionality of real networks. Another challenge is that the definition of distance has to be established according to their relational and connectivity structure, and this also requires sophisticated models.
Now, the new approach is based on the geometry of complex networks, and more specifically, on the configurational geometric model or SD model. “This model, which we have developed in previous work, describes the structure of complex networks based on fundamental principles,” says the lecturer M. Ángeles, ICREA researcher at the Department of Condensed Matter Physics of the UB.
“More specifically — he continues — , the model postulates a law of interconnection of the network elements (or nodes) that is gravitational, so nodes that are closer in a similarity space — of spherical geometry in D dimensions — and with more popularity — an extra dimension corresponding to the importance of the node — are more likely to establish connections.”
In the study, the similarity and popularity variables are combined to give rise to the hyperbolic geometry of the model, which emerges as the natural geometry representing the hierarchical architecture of complex networks.
In previous studies, the team had applied the simplest version of the one-dimensional SD model — the S1 model — to explain many typical features of real-world networks: the small-world property (the six degrees of separation), the heterogeneous distributions of the number of neighbours per node, and the high levels of transitive relationships (triangle connections that can be illustrated with the expression my friend’s friend is also my friend).
“In addition, the application of statistical inference techniques allows us to obtain real network maps in the hyperbolic plan that are congruent with the established model,” she says. “Beyond visualisation, these representations have been used in a multitude of tasks, including efficient navigation methods, the detection of self-similarity patterns, the detection of strongly interacting communities of nodes, and the implementation of a network renormalisation procedure that reveals hidden symmetries in the multi-scale organisation of complex networks and allows the production of network replicas at reduced or enlarged scales.”
Now, the team infers the dimensionality of the hyperbolic space underlying the real networks from properties that relate to the dimension of their geometry. In particular, the work measures the statistics of higher-order cycles (triangles, squares, pentagons) associated with the connections.
A methodology applicable to all complex networks
In computer science, the applied techniques are based on data that typically make definitions of similarity distance between their elements, an approach that involves the construction of graphs that are mapped onto a latent space of Euclidean features.
“Our estimates of the dimensionality of complex networks are well below our estimates based on Euclidean space, since hyperbolic space is better suited to represent the hierarchical structure of real complex networks. For example, the Internet only requires D = 7 dimensions to be mapped into the hyperbolic space of our model, whereas this name is multiplied by six and scales to D = 47 in one of the most recent techniques using Euclidean space,” says Professor Marián Boguñá.
In addition, techniques for mapping complex data usually assume a latent space, with a predetermined name of dimensions, or implement heuristic techniques to find a suitable value. Thus, the new method is based on a model that does not need the spatial mapping of the network to determine the dimension of its geometry.
In the field of network science, many methodologies use the shortest distances to study the connectivity structure of the network (shortest paths) as a metric space. However, these distances are strongly affected by the small-world property and do not provide a wide range of distance values.
“Our model uses a completely different definition of distance based on an underlying hyperbolic space, and we do not need to map the network. Our methodology is applicable to any real network or data series with complex structure and with a size that is typically thousands or tens of thousands of nodes but can reach hundreds of thousands in a reasonable computational time,” says M. Ángeles Serrano.
What is the real dimensionality of social networks and the Internet?
ocial networks and the Internet is higher (between 6 and 9) compared to networks in other domains, according to the study’s findings. However, it is still very low — 6 to 7 times lower — compared to that obtained by other methods. This reflects the fact that interactions in these systems are more complex and determined by a greater variety of factors.
On the other hand, friendship-based social networks are at the top of the dimensionality ranking. “This is an unexpected result, since one might think that friendship is a freer type of affective relationship, but our results link to the fact that homophily in human interactions is determined by a multitude of sociological factors such as age, gender, social class, beliefs, attitudes or interests,” says M. Ángeles Serrano.
In the case of the Internet, even though it is a technological network, its greater dimensionality reflects the fact that for an autonomous system, connecting does not mean only accessing the system, as one might think at first. On the contrary, many different factors influence the formation of these connections, and as a consequence, a variety of other relationships may be present (e.g., supplier-client, peer-to-peer, exchange-based peering, etc.).
“What is really surprising, both for social networks and the internet, is that our theoretical framework — which does not use any annotations about connections beyond their existence — is able to capture this multidimensional reality that is not explicit in our data,” concludes the team, which is currently working on constructing hyperbolic multidimensional maps of complex networks that are congruent with the theoretical framework established by the SD model.
Journal Reference:
- Pedro Almagro, Marián Boguñá, M. Ángeles Serrano. Detecting the ultra low dimensionality of real networks. Nature Communications, 2022; 13 (1) DOI: 10.1038/s41467-022-33685-z
Among the authors of the study are the researchers M. Ángeles Serrano and Marián Boguñá, from the Faculty of Physics and the Institute of Complex Systems of the UB (UBICS), and Pedro Almargo, from the Higher Technical School of Engineering of the University of Sevilla. The research study provides a multidimensional hyperbolic model of complex networks that reproduces its connectivity, with an ultra-low and customizable dimensionality for each specific network. This enables a better characterization of its structure — e.g. at a community scale — and the improvement of its predictive capability.
The study reveals unexpected regularities, such as the extremely low dimensions of molecular networks associated with biological tissues; the slightly higher dimensionality required by social networks and the Internet; and the discovery that brain connectomes are close to three dimensions in their automatic organisation.
Hyperbolic versus Euclidean geometry
The intrinsic geometry of data sets or complex networks is not obvious, which becomes an obstacle in determining the dimensionality of real networks. Another challenge is that the definition of distance has to be established according to their relational and connectivity structure, and this also requires sophisticated models.
Now, the new approach is based on the geometry of complex networks, and more specifically, on the configurational geometric model or SD model. “This model, which we have developed in previous work, describes the structure of complex networks based on fundamental principles,” says the lecturer M. Ángeles, ICREA researcher at the Department of Condensed Matter Physics of the UB.
“More specifically — he continues — , the model postulates a law of interconnection of the network elements (or nodes) that is gravitational, so nodes that are closer in a similarity space — of spherical geometry in D dimensions — and with more popularity — an extra dimension corresponding to the importance of the node — are more likely to establish connections.”
In the study, the similarity and popularity variables are combined to give rise to the hyperbolic geometry of the model, which emerges as the natural geometry representing the hierarchical architecture of complex networks.
In previous studies, the team had applied the simplest version of the one-dimensional SD model — the S1 model — to explain many typical features of real-world networks: the small-world property (the six degrees of separation), the heterogeneous distributions of the number of neighbours per node, and the high levels of transitive relationships (triangle connections that can be illustrated with the expression my friend’s friend is also my friend).
“In addition, the application of statistical inference techniques allows us to obtain real network maps in the hyperbolic plan that are congruent with the established model,” she says. “Beyond visualisation, these representations have been used in a multitude of tasks, including efficient navigation methods, the detection of self-similarity patterns, the detection of strongly interacting communities of nodes, and the implementation of a network renormalisation procedure that reveals hidden symmetries in the multi-scale organisation of complex networks and allows the production of network replicas at reduced or enlarged scales.”
Now, the team infers the dimensionality of the hyperbolic space underlying the real networks from properties that relate to the dimension of their geometry. In particular, the work measures the statistics of higher-order cycles (triangles, squares, pentagons) associated with the connections.
A methodology applicable to all complex networks
In computer science, the applied techniques are based on data that typically make definitions of similarity distance between their elements, an approach that involves the construction of graphs that are mapped onto a latent space of Euclidean features.
“Our estimates of the dimensionality of complex networks are well below our estimates based on Euclidean space, since hyperbolic space is better suited to represent the hierarchical structure of real complex networks. For example, the Internet only requires D = 7 dimensions to be mapped into the hyperbolic space of our model, whereas this name is multiplied by six and scales to D = 47 in one of the most recent techniques using Euclidean space,” says Professor Marián Boguñá.
In addition, techniques for mapping complex data usually assume a latent space, with a predetermined name of dimensions, or implement heuristic techniques to find a suitable value. Thus, the new method is based on a model that does not need the spatial mapping of the network to determine the dimension of its geometry.
In the field of network science, many methodologies use the shortest distances to study the connectivity structure of the network (shortest paths) as a metric space. However, these distances are strongly affected by the small-world property and do not provide a wide range of distance values.
“Our model uses a completely different definition of distance based on an underlying hyperbolic space, and we do not need to map the network. Our methodology is applicable to any real network or data series with complex structure and with a size that is typically thousands or tens of thousands of nodes but can reach hundreds of thousands in a reasonable computational time,” says M. Ángeles Serrano.
What is the real dimensionality of social networks and the Internet?
ocial networks and the Internet is higher (between 6 and 9) compared to networks in other domains, according to the study’s findings. However, it is still very low — 6 to 7 times lower — compared to that obtained by other methods. This reflects the fact that interactions in these systems are more complex and determined by a greater variety of factors.
On the other hand, friendship-based social networks are at the top of the dimensionality ranking. “This is an unexpected result, since one might think that friendship is a freer type of affective relationship, but our results link to the fact that homophily in human interactions is determined by a multitude of sociological factors such as age, gender, social class, beliefs, attitudes or interests,” says M. Ángeles Serrano.
In the case of the Internet, even though it is a technological network, its greater dimensionality reflects the fact that for an autonomous system, connecting does not mean only accessing the system, as one might think at first. On the contrary, many different factors influence the formation of these connections, and as a consequence, a variety of other relationships may be present (e.g., supplier-client, peer-to-peer, exchange-based peering, etc.).
“What is really surprising, both for social networks and the internet, is that our theoretical framework — which does not use any annotations about connections beyond their existence — is able to capture this multidimensional reality that is not explicit in our data,” concludes the team, which is currently working on constructing hyperbolic multidimensional maps of complex networks that are congruent with the theoretical framework established by the SD model.
Source link
The post Unveiling the dimensionality of complex networks through hyperbolic geometry first appeared on .
)
[5] => Array
(
[title] => Good sleep can increase women’s work ambitions
[link] => https://coolnspicy.com/science/good-sleep-can-increase-womens-work-ambitions/
[dc] => Array
(
[creator] => Michael Steiner
)
[pubdate] => Thu, 03 Nov 2022 08:06:03 +0000
[category] => scienceambitionsincreasesleepwomens
[guid] => https://coolnspicy.com/?p=289
[description] => Journal Reference: Leah D. Sheppard, Teng Iat Loi, Julie A. Kmec. Too Tired to Lean In? Sleep Quality Impacts Women’s Daily Intentions to Pursue Workplace Status. Sex Roles, 2022; DOI: 10.1007/s11199-022-01321-1 The researchers discovered this finding in a two-week-long survey study of 135 workers in the U.S. Each day the participants first noted how well ... Read more
The post Good sleep can increase women’s work ambitions first appeared on .
[content] => Array
(
[encoded] =>
Journal Reference:
- Leah D. Sheppard, Teng Iat Loi, Julie A. Kmec. Too Tired to Lean In? Sleep Quality Impacts Women’s Daily Intentions to Pursue Workplace Status. Sex Roles, 2022; DOI: 10.1007/s11199-022-01321-1
The researchers discovered this finding in a two-week-long survey study of 135 workers in the U.S. Each day the participants first noted how well they had slept and the quality of their current mood, and then later in the day how they felt about striving for more status and responsibility at work.
“When women are getting a good night’s sleep and their mood is boosted, they are more likely to be oriented in their daily intentions toward achieving status and responsibility at work,” said lead author Leah Sheppard, an associate professor in WSU’s Carson College of Business. “If their sleep is poor and reduces their positive mood, then we saw that they were less oriented toward those goals.”
For the study published in the journal Sex Roles, Sheppard and co-authors Julie Kmec of WSU and Teng Iat Loi of University of Minnesota-Duluth surveyed full-time employees twice a day for two consecutive work weeks for a total of more than 2,200 observations. The participants answered questions about their previous night’s sleep and current mood around noon every day and in the evenings answered questions about their intentions to pursue more responsibility, status, and influence at work.
Both men and women reported good and bad sleep quality over the course of the study, notably with no gender difference in reported sleep quality. However, women more often reported lowered intentions to pursue more status at work on days following a night of poor sleep.
The researchers can only speculate about exactly why sleep’s impact on mood effects women’s aspirations and not men’s, but they suspect it may have to do with gender differences in emotion regulation as well as societal expectations — or some combination of these forces.
Neuroscience research has shown that women tend to experience greater emotional re-activity and less emotion regulation than men, and this can be reinforced by cultural stereotypes of women as more emotional. At the same time, stereotypes of men as being more ambitious than women likely add more pressure for them to scale the corporate ladder, so perhaps poor sleep quality would be less likely to deter men from their work aspirations.
These findings hold some good news for women who want to advance their careers, though, Sheppard said. For instance, they might take some practical steps to improve work aspirations, ranging from practicing meditation to help with both sleep and emotion regulation to putting better boundaries on work hours — and of course, simply striving to get better sleep.
“It’s important to be able to connect aspirations to something happening outside the work environment that is controllable,” she said. “There are lots of things that anyone can do to have a better night’s sleep and regulate mood in general.”
Journal Reference:
- Leah D. Sheppard, Teng Iat Loi, Julie A. Kmec. Too Tired to Lean In? Sleep Quality Impacts Women’s Daily Intentions to Pursue Workplace Status. Sex Roles, 2022; DOI: 10.1007/s11199-022-01321-1
The researchers discovered this finding in a two-week-long survey study of 135 workers in the U.S. Each day the participants first noted how well they had slept and the quality of their current mood, and then later in the day how they felt about striving for more status and responsibility at work.
“When women are getting a good night’s sleep and their mood is boosted, they are more likely to be oriented in their daily intentions toward achieving status and responsibility at work,” said lead author Leah Sheppard, an associate professor in WSU’s Carson College of Business. “If their sleep is poor and reduces their positive mood, then we saw that they were less oriented toward those goals.”
For the study published in the journal Sex Roles, Sheppard and co-authors Julie Kmec of WSU and Teng Iat Loi of University of Minnesota-Duluth surveyed full-time employees twice a day for two consecutive work weeks for a total of more than 2,200 observations. The participants answered questions about their previous night’s sleep and current mood around noon every day and in the evenings answered questions about their intentions to pursue more responsibility, status, and influence at work.
Both men and women reported good and bad sleep quality over the course of the study, notably with no gender difference in reported sleep quality. However, women more often reported lowered intentions to pursue more status at work on days following a night of poor sleep.
The researchers can only speculate about exactly why sleep’s impact on mood effects women’s aspirations and not men’s, but they suspect it may have to do with gender differences in emotion regulation as well as societal expectations — or some combination of these forces.
Neuroscience research has shown that women tend to experience greater emotional re-activity and less emotion regulation than men, and this can be reinforced by cultural stereotypes of women as more emotional. At the same time, stereotypes of men as being more ambitious than women likely add more pressure for them to scale the corporate ladder, so perhaps poor sleep quality would be less likely to deter men from their work aspirations.
These findings hold some good news for women who want to advance their careers, though, Sheppard said. For instance, they might take some practical steps to improve work aspirations, ranging from practicing meditation to help with both sleep and emotion regulation to putting better boundaries on work hours — and of course, simply striving to get better sleep.
“It’s important to be able to connect aspirations to something happening outside the work environment that is controllable,” she said. “There are lots of things that anyone can do to have a better night’s sleep and regulate mood in general.”
Source link
The post Good sleep can increase women’s work ambitions first appeared on .
)
[summary] => Journal Reference: Leah D. Sheppard, Teng Iat Loi, Julie A. Kmec. Too Tired to Lean In? Sleep Quality Impacts Women’s Daily Intentions to Pursue Workplace Status. Sex Roles, 2022; DOI: 10.1007/s11199-022-01321-1 The researchers discovered this finding in a two-week-long survey study of 135 workers in the U.S. Each day the participants first noted how well ... Read more
The post Good sleep can increase women’s work ambitions first appeared on .
[atom_content] =>
Journal Reference:
- Leah D. Sheppard, Teng Iat Loi, Julie A. Kmec. Too Tired to Lean In? Sleep Quality Impacts Women’s Daily Intentions to Pursue Workplace Status. Sex Roles, 2022; DOI: 10.1007/s11199-022-01321-1
The researchers discovered this finding in a two-week-long survey study of 135 workers in the U.S. Each day the participants first noted how well they had slept and the quality of their current mood, and then later in the day how they felt about striving for more status and responsibility at work.
“When women are getting a good night’s sleep and their mood is boosted, they are more likely to be oriented in their daily intentions toward achieving status and responsibility at work,” said lead author Leah Sheppard, an associate professor in WSU’s Carson College of Business. “If their sleep is poor and reduces their positive mood, then we saw that they were less oriented toward those goals.”
For the study published in the journal Sex Roles, Sheppard and co-authors Julie Kmec of WSU and Teng Iat Loi of University of Minnesota-Duluth surveyed full-time employees twice a day for two consecutive work weeks for a total of more than 2,200 observations. The participants answered questions about their previous night’s sleep and current mood around noon every day and in the evenings answered questions about their intentions to pursue more responsibility, status, and influence at work.
Both men and women reported good and bad sleep quality over the course of the study, notably with no gender difference in reported sleep quality. However, women more often reported lowered intentions to pursue more status at work on days following a night of poor sleep.
The researchers can only speculate about exactly why sleep’s impact on mood effects women’s aspirations and not men’s, but they suspect it may have to do with gender differences in emotion regulation as well as societal expectations — or some combination of these forces.
Neuroscience research has shown that women tend to experience greater emotional re-activity and less emotion regulation than men, and this can be reinforced by cultural stereotypes of women as more emotional. At the same time, stereotypes of men as being more ambitious than women likely add more pressure for them to scale the corporate ladder, so perhaps poor sleep quality would be less likely to deter men from their work aspirations.
These findings hold some good news for women who want to advance their careers, though, Sheppard said. For instance, they might take some practical steps to improve work aspirations, ranging from practicing meditation to help with both sleep and emotion regulation to putting better boundaries on work hours — and of course, simply striving to get better sleep.
“It’s important to be able to connect aspirations to something happening outside the work environment that is controllable,” she said. “There are lots of things that anyone can do to have a better night’s sleep and regulate mood in general.”
Journal Reference:
- Leah D. Sheppard, Teng Iat Loi, Julie A. Kmec. Too Tired to Lean In? Sleep Quality Impacts Women’s Daily Intentions to Pursue Workplace Status. Sex Roles, 2022; DOI: 10.1007/s11199-022-01321-1
The researchers discovered this finding in a two-week-long survey study of 135 workers in the U.S. Each day the participants first noted how well they had slept and the quality of their current mood, and then later in the day how they felt about striving for more status and responsibility at work.
“When women are getting a good night’s sleep and their mood is boosted, they are more likely to be oriented in their daily intentions toward achieving status and responsibility at work,” said lead author Leah Sheppard, an associate professor in WSU’s Carson College of Business. “If their sleep is poor and reduces their positive mood, then we saw that they were less oriented toward those goals.”
For the study published in the journal Sex Roles, Sheppard and co-authors Julie Kmec of WSU and Teng Iat Loi of University of Minnesota-Duluth surveyed full-time employees twice a day for two consecutive work weeks for a total of more than 2,200 observations. The participants answered questions about their previous night’s sleep and current mood around noon every day and in the evenings answered questions about their intentions to pursue more responsibility, status, and influence at work.
Both men and women reported good and bad sleep quality over the course of the study, notably with no gender difference in reported sleep quality. However, women more often reported lowered intentions to pursue more status at work on days following a night of poor sleep.
The researchers can only speculate about exactly why sleep’s impact on mood effects women’s aspirations and not men’s, but they suspect it may have to do with gender differences in emotion regulation as well as societal expectations — or some combination of these forces.
Neuroscience research has shown that women tend to experience greater emotional re-activity and less emotion regulation than men, and this can be reinforced by cultural stereotypes of women as more emotional. At the same time, stereotypes of men as being more ambitious than women likely add more pressure for them to scale the corporate ladder, so perhaps poor sleep quality would be less likely to deter men from their work aspirations.
These findings hold some good news for women who want to advance their careers, though, Sheppard said. For instance, they might take some practical steps to improve work aspirations, ranging from practicing meditation to help with both sleep and emotion regulation to putting better boundaries on work hours — and of course, simply striving to get better sleep.
“It’s important to be able to connect aspirations to something happening outside the work environment that is controllable,” she said. “There are lots of things that anyone can do to have a better night’s sleep and regulate mood in general.”
Source link
The post Good sleep can increase women’s work ambitions first appeared on .
)
[6] => Array
(
[title] => Machine learning facilitates ‘turbulence monitoring’ in fusion reactors
[link] => https://coolnspicy.com/science/machine-learning-facilitates-turbulence-monitoring-in-fusion-reactors/
[dc] => Array
(
[creator] => Michael Steiner
)
[pubdate] => Thu, 03 Nov 2022 08:04:08 +0000
[category] => sciencefacilitatesfusionlearningMachinemonitoringreactorsturbulence
[guid] => https://coolnspicy.com/?p=286
[description] => Journal Reference: Woonghee Han, Randall A. Pietersen, Rafael Villamor-Lora, Matthew Beveridge, Nicola Offeddu, Theodore Golfinopoulos, Christian Theiler, James L. Terry, Earl S. Marmar, Iddo Drori. Tracking blobs in the turbulent edge plasma of a tokamak fusion device. Scientific Reports, 2022; 12 (1) DOI: 10.1038/s41598-022-21671-w A multidisciplinary team of researchers is now bringing tools and insights ... Read more
The post Machine learning facilitates ‘turbulence monitoring’ in fusion reactors first appeared on .
[content] => Array
(
[encoded] =>
Journal Reference:
- Woonghee Han, Randall A. Pietersen, Rafael Villamor-Lora, Matthew Beveridge, Nicola Offeddu, Theodore Golfinopoulos, Christian Theiler, James L. Terry, Earl S. Marmar, Iddo Drori. Tracking blobs in the turbulent edge plasma of a tokamak fusion device. Scientific Reports, 2022; 12 (1) DOI: 10.1038/s41598-022-21671-w
A multidisciplinary team of researchers is now bringing tools and insights from machine learning to aid this effort. Scientists from MIT and elsewhere have used computer-vision models to identify and track turbulent structures that appear under the conditions needed to facilitate fusion reactions.
Monitoring the formation and movements of these structures, called filaments or “blobs,” is important for understanding the heat and particle flows exiting from the reacting fuel, which ultimately determines the engineering requirements for the reactor walls to meet those flows. However, scientists typically study blobs using averaging techniques, which trade details of individual structures in favor of aggregate statistics. Individual blob information must be tracked by marking them manually in video data.
The researchers built a synthetic video dataset of plasma turbulence to make this process more effective and efficient. They used it to train four computer vision models, each of which identifies and tracks blobs. They trained the models to pinpoint blobs in the same ways that humans would.
When the researchers tested the trained models using real video clips, the models could identify blobs with high accuracy — more than 80 percent in some cases. The models were also able to effectively estimate the size of blobs and the speeds at which they moved.
Because millions of video frames are captured during just one fusion experiment, using machine-learning models to track blobs could give scientists much more detailed information.
“Before, we could get a macroscopic picture of what these structures are doing on average. Now, we have a microscope and the computational power to analyze one event at a time. If we take a step back, what this reveals is the power available from these machine-learning techniques, and ways to use these computational resources to make progress,” says Theodore Golfinopoulos, a research scientist at the MIT Plasma Science and Fusion Center and co-author of a paper detailing these approaches.
His fellow co-authors include lead author Woonghee “Harry” Han, a physics PhD candidate; senior author Iddo Drori, a visiting professor in the Computer Science and Artificial Intelligence Laboratory (CSAIL), faculty associate professor at Boston University, and adjunct at Columbia University; as well as others from the MIT Plasma Science and Fusion Center, the MIT Department of Civil and Environmental Engineering, and the Swiss Federal Institute of Technology at Lausanne in Switzerland. The research appears today in Nature Scientific Reports.
Heating things up
For more than 70 years, scientists have sought to use controlled thermonuclear fusion reactions to develop an energy source. To reach the conditions necessary for a fusion reaction, fuel must be heated to temperatures above 100 million degrees Celsius. (The core of the sun is about 15 million degrees Celsius.)
A common method for containing this super-hot fuel, called plasma, is to use a tokamak. These devices utilize extremely powerful magnetic fields to hold the plasma in place and control the interaction between the exhaust heat from the plasma and the reactor walls.
However, blobs appear like filaments falling out of the plasma at the very edge, between the plasma and the reactor walls. These random, turbulent structures affect how energy flows between the plasma and the reactor.
“Knowing what the blobs are doing strongly constrains the engineering performance that your tokamak power plant needs at the edge,” adds Golfinopoulos.
Researchers use a unique imaging technique to capture video of the plasma’s turbulent edge during experiments. An experimental campaign may last months; a typical day will produce about 30 seconds of data, corresponding to roughly 60 million video frames, with thousands of blobs appearing each second. This makes it impossible to track all blobs manually, so researchers rely on average sampling techniques that only provide broad characteristics of blob size, speed, and frequency.
“On the other hand, machine learning provides a solution to this by blob-by-blob tracking for every frame, not just average quantities. This gives us much more knowledge about what is happening at the boundary of the plasma,” Han says.
He and his co-authors took four well-established computer vision models, which are commonly used for applications like autonomous driving, and trained them to tackle this problem.
Simulating blobs
To train these models, they created a vast dataset of synthetic video clips that captured the blobs’ random and unpredictable nature.
“Sometimes they change direction or speed, sometimes multiple blobs merge, or they split apart. These kinds of events were not considered before with traditional approaches, but we could freely simulate those behaviors in the synthetic data,” Han says.
Creating synthetic data also allowed them to label each blob, which made the training process more effective, Drori adds.
Using these synthetic data, they trained the models to draw boundaries around blobs, teaching them to closely mimic what a human scientist would draw.
Then they tested the models using real video data from experiments. First, they measured how closely the boundaries the models drew matched up with actual blob contours.
But they also wanted to see if the models predicted objects that humans would identify. They asked three human experts to pinpoint the centers of blobs in video frames and checked to see if the models predicted blobs in those same locations.
The models were able to draw accurate blob boundaries, overlapping with brightness contours which are considered ground-truth, about 80 percent of the time. Their evaluations were similar to those of human experts, and successfully predicted the theory-defined regime of the blob, which agrees with the results from a traditional method.
Now that they have shown the success of using synthetic data and computer vision models for tracking blobs, the researchers plan to apply these techniques to other problems in fusion research, such as estimating particle transport at the boundary of a plasma, Han says.
They also made the dataset and models publicly available, and look forward to seeing how other research groups apply these tools to study the dynamics of blobs, says Drori.
“Prior to this, there was a barrier to entry that mostly the only people working on this problem were plasma physicists, who had the datasets and were using their methods. There is a huge machine-learning and computer-vision community. One goal of this work is to encourage participation in fusion research from the broader machine-learning community toward the broader goal of helping solve the critical problem of climate change,” he adds.
This research is supported, in part, by the U.S. Department of Energy and the Swiss National Science Foundation.
Journal Reference:
- Woonghee Han, Randall A. Pietersen, Rafael Villamor-Lora, Matthew Beveridge, Nicola Offeddu, Theodore Golfinopoulos, Christian Theiler, James L. Terry, Earl S. Marmar, Iddo Drori. Tracking blobs in the turbulent edge plasma of a tokamak fusion device. Scientific Reports, 2022; 12 (1) DOI: 10.1038/s41598-022-21671-w
A multidisciplinary team of researchers is now bringing tools and insights from machine learning to aid this effort. Scientists from MIT and elsewhere have used computer-vision models to identify and track turbulent structures that appear under the conditions needed to facilitate fusion reactions.
Monitoring the formation and movements of these structures, called filaments or “blobs,” is important for understanding the heat and particle flows exiting from the reacting fuel, which ultimately determines the engineering requirements for the reactor walls to meet those flows. However, scientists typically study blobs using averaging techniques, which trade details of individual structures in favor of aggregate statistics. Individual blob information must be tracked by marking them manually in video data.
The researchers built a synthetic video dataset of plasma turbulence to make this process more effective and efficient. They used it to train four computer vision models, each of which identifies and tracks blobs. They trained the models to pinpoint blobs in the same ways that humans would.
When the researchers tested the trained models using real video clips, the models could identify blobs with high accuracy — more than 80 percent in some cases. The models were also able to effectively estimate the size of blobs and the speeds at which they moved.
Because millions of video frames are captured during just one fusion experiment, using machine-learning models to track blobs could give scientists much more detailed information.
“Before, we could get a macroscopic picture of what these structures are doing on average. Now, we have a microscope and the computational power to analyze one event at a time. If we take a step back, what this reveals is the power available from these machine-learning techniques, and ways to use these computational resources to make progress,” says Theodore Golfinopoulos, a research scientist at the MIT Plasma Science and Fusion Center and co-author of a paper detailing these approaches.
His fellow co-authors include lead author Woonghee “Harry” Han, a physics PhD candidate; senior author Iddo Drori, a visiting professor in the Computer Science and Artificial Intelligence Laboratory (CSAIL), faculty associate professor at Boston University, and adjunct at Columbia University; as well as others from the MIT Plasma Science and Fusion Center, the MIT Department of Civil and Environmental Engineering, and the Swiss Federal Institute of Technology at Lausanne in Switzerland. The research appears today in Nature Scientific Reports.
Heating things up
For more than 70 years, scientists have sought to use controlled thermonuclear fusion reactions to develop an energy source. To reach the conditions necessary for a fusion reaction, fuel must be heated to temperatures above 100 million degrees Celsius. (The core of the sun is about 15 million degrees Celsius.)
A common method for containing this super-hot fuel, called plasma, is to use a tokamak. These devices utilize extremely powerful magnetic fields to hold the plasma in place and control the interaction between the exhaust heat from the plasma and the reactor walls.
However, blobs appear like filaments falling out of the plasma at the very edge, between the plasma and the reactor walls. These random, turbulent structures affect how energy flows between the plasma and the reactor.
“Knowing what the blobs are doing strongly constrains the engineering performance that your tokamak power plant needs at the edge,” adds Golfinopoulos.
Researchers use a unique imaging technique to capture video of the plasma’s turbulent edge during experiments. An experimental campaign may last months; a typical day will produce about 30 seconds of data, corresponding to roughly 60 million video frames, with thousands of blobs appearing each second. This makes it impossible to track all blobs manually, so researchers rely on average sampling techniques that only provide broad characteristics of blob size, speed, and frequency.
“On the other hand, machine learning provides a solution to this by blob-by-blob tracking for every frame, not just average quantities. This gives us much more knowledge about what is happening at the boundary of the plasma,” Han says.
He and his co-authors took four well-established computer vision models, which are commonly used for applications like autonomous driving, and trained them to tackle this problem.
Simulating blobs
To train these models, they created a vast dataset of synthetic video clips that captured the blobs’ random and unpredictable nature.
“Sometimes they change direction or speed, sometimes multiple blobs merge, or they split apart. These kinds of events were not considered before with traditional approaches, but we could freely simulate those behaviors in the synthetic data,” Han says.
Creating synthetic data also allowed them to label each blob, which made the training process more effective, Drori adds.
Using these synthetic data, they trained the models to draw boundaries around blobs, teaching them to closely mimic what a human scientist would draw.
Then they tested the models using real video data from experiments. First, they measured how closely the boundaries the models drew matched up with actual blob contours.
But they also wanted to see if the models predicted objects that humans would identify. They asked three human experts to pinpoint the centers of blobs in video frames and checked to see if the models predicted blobs in those same locations.
The models were able to draw accurate blob boundaries, overlapping with brightness contours which are considered ground-truth, about 80 percent of the time. Their evaluations were similar to those of human experts, and successfully predicted the theory-defined regime of the blob, which agrees with the results from a traditional method.
Now that they have shown the success of using synthetic data and computer vision models for tracking blobs, the researchers plan to apply these techniques to other problems in fusion research, such as estimating particle transport at the boundary of a plasma, Han says.
They also made the dataset and models publicly available, and look forward to seeing how other research groups apply these tools to study the dynamics of blobs, says Drori.
“Prior to this, there was a barrier to entry that mostly the only people working on this problem were plasma physicists, who had the datasets and were using their methods. There is a huge machine-learning and computer-vision community. One goal of this work is to encourage participation in fusion research from the broader machine-learning community toward the broader goal of helping solve the critical problem of climate change,” he adds.
This research is supported, in part, by the U.S. Department of Energy and the Swiss National Science Foundation.
Source by
The post Machine learning facilitates ‘turbulence monitoring’ in fusion reactors first appeared on .
)
[summary] => Journal Reference: Woonghee Han, Randall A. Pietersen, Rafael Villamor-Lora, Matthew Beveridge, Nicola Offeddu, Theodore Golfinopoulos, Christian Theiler, James L. Terry, Earl S. Marmar, Iddo Drori. Tracking blobs in the turbulent edge plasma of a tokamak fusion device. Scientific Reports, 2022; 12 (1) DOI: 10.1038/s41598-022-21671-w A multidisciplinary team of researchers is now bringing tools and insights ... Read more
The post Machine learning facilitates ‘turbulence monitoring’ in fusion reactors first appeared on .
[atom_content] =>
Journal Reference:
- Woonghee Han, Randall A. Pietersen, Rafael Villamor-Lora, Matthew Beveridge, Nicola Offeddu, Theodore Golfinopoulos, Christian Theiler, James L. Terry, Earl S. Marmar, Iddo Drori. Tracking blobs in the turbulent edge plasma of a tokamak fusion device. Scientific Reports, 2022; 12 (1) DOI: 10.1038/s41598-022-21671-w
A multidisciplinary team of researchers is now bringing tools and insights from machine learning to aid this effort. Scientists from MIT and elsewhere have used computer-vision models to identify and track turbulent structures that appear under the conditions needed to facilitate fusion reactions.
Monitoring the formation and movements of these structures, called filaments or “blobs,” is important for understanding the heat and particle flows exiting from the reacting fuel, which ultimately determines the engineering requirements for the reactor walls to meet those flows. However, scientists typically study blobs using averaging techniques, which trade details of individual structures in favor of aggregate statistics. Individual blob information must be tracked by marking them manually in video data.
The researchers built a synthetic video dataset of plasma turbulence to make this process more effective and efficient. They used it to train four computer vision models, each of which identifies and tracks blobs. They trained the models to pinpoint blobs in the same ways that humans would.
When the researchers tested the trained models using real video clips, the models could identify blobs with high accuracy — more than 80 percent in some cases. The models were also able to effectively estimate the size of blobs and the speeds at which they moved.
Because millions of video frames are captured during just one fusion experiment, using machine-learning models to track blobs could give scientists much more detailed information.
“Before, we could get a macroscopic picture of what these structures are doing on average. Now, we have a microscope and the computational power to analyze one event at a time. If we take a step back, what this reveals is the power available from these machine-learning techniques, and ways to use these computational resources to make progress,” says Theodore Golfinopoulos, a research scientist at the MIT Plasma Science and Fusion Center and co-author of a paper detailing these approaches.
His fellow co-authors include lead author Woonghee “Harry” Han, a physics PhD candidate; senior author Iddo Drori, a visiting professor in the Computer Science and Artificial Intelligence Laboratory (CSAIL), faculty associate professor at Boston University, and adjunct at Columbia University; as well as others from the MIT Plasma Science and Fusion Center, the MIT Department of Civil and Environmental Engineering, and the Swiss Federal Institute of Technology at Lausanne in Switzerland. The research appears today in Nature Scientific Reports.
Heating things up
For more than 70 years, scientists have sought to use controlled thermonuclear fusion reactions to develop an energy source. To reach the conditions necessary for a fusion reaction, fuel must be heated to temperatures above 100 million degrees Celsius. (The core of the sun is about 15 million degrees Celsius.)
A common method for containing this super-hot fuel, called plasma, is to use a tokamak. These devices utilize extremely powerful magnetic fields to hold the plasma in place and control the interaction between the exhaust heat from the plasma and the reactor walls.
However, blobs appear like filaments falling out of the plasma at the very edge, between the plasma and the reactor walls. These random, turbulent structures affect how energy flows between the plasma and the reactor.
“Knowing what the blobs are doing strongly constrains the engineering performance that your tokamak power plant needs at the edge,” adds Golfinopoulos.
Researchers use a unique imaging technique to capture video of the plasma’s turbulent edge during experiments. An experimental campaign may last months; a typical day will produce about 30 seconds of data, corresponding to roughly 60 million video frames, with thousands of blobs appearing each second. This makes it impossible to track all blobs manually, so researchers rely on average sampling techniques that only provide broad characteristics of blob size, speed, and frequency.
“On the other hand, machine learning provides a solution to this by blob-by-blob tracking for every frame, not just average quantities. This gives us much more knowledge about what is happening at the boundary of the plasma,” Han says.
He and his co-authors took four well-established computer vision models, which are commonly used for applications like autonomous driving, and trained them to tackle this problem.
Simulating blobs
To train these models, they created a vast dataset of synthetic video clips that captured the blobs’ random and unpredictable nature.
“Sometimes they change direction or speed, sometimes multiple blobs merge, or they split apart. These kinds of events were not considered before with traditional approaches, but we could freely simulate those behaviors in the synthetic data,” Han says.
Creating synthetic data also allowed them to label each blob, which made the training process more effective, Drori adds.
Using these synthetic data, they trained the models to draw boundaries around blobs, teaching them to closely mimic what a human scientist would draw.
Then they tested the models using real video data from experiments. First, they measured how closely the boundaries the models drew matched up with actual blob contours.
But they also wanted to see if the models predicted objects that humans would identify. They asked three human experts to pinpoint the centers of blobs in video frames and checked to see if the models predicted blobs in those same locations.
The models were able to draw accurate blob boundaries, overlapping with brightness contours which are considered ground-truth, about 80 percent of the time. Their evaluations were similar to those of human experts, and successfully predicted the theory-defined regime of the blob, which agrees with the results from a traditional method.
Now that they have shown the success of using synthetic data and computer vision models for tracking blobs, the researchers plan to apply these techniques to other problems in fusion research, such as estimating particle transport at the boundary of a plasma, Han says.
They also made the dataset and models publicly available, and look forward to seeing how other research groups apply these tools to study the dynamics of blobs, says Drori.
“Prior to this, there was a barrier to entry that mostly the only people working on this problem were plasma physicists, who had the datasets and were using their methods. There is a huge machine-learning and computer-vision community. One goal of this work is to encourage participation in fusion research from the broader machine-learning community toward the broader goal of helping solve the critical problem of climate change,” he adds.
This research is supported, in part, by the U.S. Department of Energy and the Swiss National Science Foundation.
Journal Reference:
- Woonghee Han, Randall A. Pietersen, Rafael Villamor-Lora, Matthew Beveridge, Nicola Offeddu, Theodore Golfinopoulos, Christian Theiler, James L. Terry, Earl S. Marmar, Iddo Drori. Tracking blobs in the turbulent edge plasma of a tokamak fusion device. Scientific Reports, 2022; 12 (1) DOI: 10.1038/s41598-022-21671-w
A multidisciplinary team of researchers is now bringing tools and insights from machine learning to aid this effort. Scientists from MIT and elsewhere have used computer-vision models to identify and track turbulent structures that appear under the conditions needed to facilitate fusion reactions.
Monitoring the formation and movements of these structures, called filaments or “blobs,” is important for understanding the heat and particle flows exiting from the reacting fuel, which ultimately determines the engineering requirements for the reactor walls to meet those flows. However, scientists typically study blobs using averaging techniques, which trade details of individual structures in favor of aggregate statistics. Individual blob information must be tracked by marking them manually in video data.
The researchers built a synthetic video dataset of plasma turbulence to make this process more effective and efficient. They used it to train four computer vision models, each of which identifies and tracks blobs. They trained the models to pinpoint blobs in the same ways that humans would.
When the researchers tested the trained models using real video clips, the models could identify blobs with high accuracy — more than 80 percent in some cases. The models were also able to effectively estimate the size of blobs and the speeds at which they moved.
Because millions of video frames are captured during just one fusion experiment, using machine-learning models to track blobs could give scientists much more detailed information.
“Before, we could get a macroscopic picture of what these structures are doing on average. Now, we have a microscope and the computational power to analyze one event at a time. If we take a step back, what this reveals is the power available from these machine-learning techniques, and ways to use these computational resources to make progress,” says Theodore Golfinopoulos, a research scientist at the MIT Plasma Science and Fusion Center and co-author of a paper detailing these approaches.
His fellow co-authors include lead author Woonghee “Harry” Han, a physics PhD candidate; senior author Iddo Drori, a visiting professor in the Computer Science and Artificial Intelligence Laboratory (CSAIL), faculty associate professor at Boston University, and adjunct at Columbia University; as well as others from the MIT Plasma Science and Fusion Center, the MIT Department of Civil and Environmental Engineering, and the Swiss Federal Institute of Technology at Lausanne in Switzerland. The research appears today in Nature Scientific Reports.
Heating things up
For more than 70 years, scientists have sought to use controlled thermonuclear fusion reactions to develop an energy source. To reach the conditions necessary for a fusion reaction, fuel must be heated to temperatures above 100 million degrees Celsius. (The core of the sun is about 15 million degrees Celsius.)
A common method for containing this super-hot fuel, called plasma, is to use a tokamak. These devices utilize extremely powerful magnetic fields to hold the plasma in place and control the interaction between the exhaust heat from the plasma and the reactor walls.
However, blobs appear like filaments falling out of the plasma at the very edge, between the plasma and the reactor walls. These random, turbulent structures affect how energy flows between the plasma and the reactor.
“Knowing what the blobs are doing strongly constrains the engineering performance that your tokamak power plant needs at the edge,” adds Golfinopoulos.
Researchers use a unique imaging technique to capture video of the plasma’s turbulent edge during experiments. An experimental campaign may last months; a typical day will produce about 30 seconds of data, corresponding to roughly 60 million video frames, with thousands of blobs appearing each second. This makes it impossible to track all blobs manually, so researchers rely on average sampling techniques that only provide broad characteristics of blob size, speed, and frequency.
“On the other hand, machine learning provides a solution to this by blob-by-blob tracking for every frame, not just average quantities. This gives us much more knowledge about what is happening at the boundary of the plasma,” Han says.
He and his co-authors took four well-established computer vision models, which are commonly used for applications like autonomous driving, and trained them to tackle this problem.
Simulating blobs
To train these models, they created a vast dataset of synthetic video clips that captured the blobs’ random and unpredictable nature.
“Sometimes they change direction or speed, sometimes multiple blobs merge, or they split apart. These kinds of events were not considered before with traditional approaches, but we could freely simulate those behaviors in the synthetic data,” Han says.
Creating synthetic data also allowed them to label each blob, which made the training process more effective, Drori adds.
Using these synthetic data, they trained the models to draw boundaries around blobs, teaching them to closely mimic what a human scientist would draw.
Then they tested the models using real video data from experiments. First, they measured how closely the boundaries the models drew matched up with actual blob contours.
But they also wanted to see if the models predicted objects that humans would identify. They asked three human experts to pinpoint the centers of blobs in video frames and checked to see if the models predicted blobs in those same locations.
The models were able to draw accurate blob boundaries, overlapping with brightness contours which are considered ground-truth, about 80 percent of the time. Their evaluations were similar to those of human experts, and successfully predicted the theory-defined regime of the blob, which agrees with the results from a traditional method.
Now that they have shown the success of using synthetic data and computer vision models for tracking blobs, the researchers plan to apply these techniques to other problems in fusion research, such as estimating particle transport at the boundary of a plasma, Han says.
They also made the dataset and models publicly available, and look forward to seeing how other research groups apply these tools to study the dynamics of blobs, says Drori.
“Prior to this, there was a barrier to entry that mostly the only people working on this problem were plasma physicists, who had the datasets and were using their methods. There is a huge machine-learning and computer-vision community. One goal of this work is to encourage participation in fusion research from the broader machine-learning community toward the broader goal of helping solve the critical problem of climate change,” he adds.
This research is supported, in part, by the U.S. Department of Energy and the Swiss National Science Foundation.
Source by
The post Machine learning facilitates ‘turbulence monitoring’ in fusion reactors first appeared on .
)
[7] => Array
(
[title] => A common dietary fiber promotes allergy
[link] => https://coolnspicy.com/science/a-common-dietary-fiber-promotes-allergy/
[dc] => Array
(
[creator] => Michael Steiner
)
[pubdate] => Thu, 03 Nov 2022 08:02:04 +0000
[category] => scienceallergycommondietaryfiberpromotes
[guid] => https://coolnspicy.com/?p=284
[description] => Journal Reference: Mohammad Arifuzzaman, Tae Hyung Won, Ting-Ting Li, Hiroshi Yano, Sreehaas Digumarthi, Andrea F. Heras, Wen Zhang, Christopher N. Parkhurst, Sanchita Kashyap, Wen-Bing Jin, Gregory Garbès Putzel, Amy M. Tsou, Coco Chu, Qianru Wei, Alex Grier, Randy Longman, Gregory Sonnenberg, Ellen Scherl, Robbyn Sockolow, Dana Lukin, Robert Battat, Thomas Ciecierega, Aliza Solomon, Elaine Barfield, ... Read more
The post A common dietary fiber promotes allergy first appeared on .
[content] => Array
(
[encoded] =>
Journal Reference:
- Mohammad Arifuzzaman, Tae Hyung Won, Ting-Ting Li, Hiroshi Yano, Sreehaas Digumarthi, Andrea F. Heras, Wen Zhang, Christopher N. Parkhurst, Sanchita Kashyap, Wen-Bing Jin, Gregory Garbès Putzel, Amy M. Tsou, Coco Chu, Qianru Wei, Alex Grier, Randy Longman, Gregory Sonnenberg, Ellen Scherl, Robbyn Sockolow, Dana Lukin, Robert Battat, Thomas Ciecierega, Aliza Solomon, Elaine Barfield, Kimberley Chien, Johanna Ferreira, Jasmin Williams, Shaira Khan, Peik Sean Chong, Samah Mozumder, Lance Chou, Wenqing Zhou, Anees Ahmed, Connie Zhong, Ann Joseph, Joseph Gladstone, Samantha Jensen, Stefan Worgall, Chun-Jun Guo, Frank C. Schroeder, David Artis. Inulin fibre promotes microbiota-derived bile acids and type 2 inflammation. Nature, 2022; DOI: 10.1038/s41586-022-05380-y
The study, published Nov. 2 in Nature, found that dietary inulin fiber alters the metabolism of certain gut bacteria, which in turn triggers what scientists call type 2 inflammation in the gut and lungs. This type of inflammation is thought to have evolved in mammals chiefly to defend against parasitic worm (“helminth”) infections, and is also part of normal wound-healing, although its inappropriate activation underlies allergies, asthma and other inflammatory diseases.
“There’s a lot to think about here, but, in general, these findings broaden our understanding of the relationship between diet, immunity, and the normally beneficial microorganisms that constitute our microbiota and colonize our bodies,” said study co-senior author Dr. David Artis, director of the Friedman Center for Nutrition and Inflammation and the Michael Kors Professor of Immunology at Weill Cornell Medicine.
The study’s scientific participants reflect the Friedman Center’s highly cross-collaborative research mission, drawing on expertise in bacterial genetics, biochemistry and immunology at Weill Cornell Medicine in New York City and Cornell’s Ithaca campus. Dr. Chun-Jun Guo, assistant professor of immunology in medicine at Weill Cornell Medicine, and Dr. Frank Schroeder, professor at the Boyce Thompson Institute and in the Department of Chemistry and Chemical Biology in the College of Arts and Sciences on Cornell’s Ithaca campus teamed up with the Artis laboratory to gain a detailed understanding of how an important dietary component affects the microbiome and the immune response. The study’s first author is Dr. Mohammad Arifuzzaman, a postdoctoral researcher in the Artis laboratory.Dr. Artis is also director of the Jill Roberts Institute for Inflammatory Bowel Disease at Weill Cornell Medicine.
Small amounts of inulin are present in a wide variety of fruits and vegetables, including bananas, asparagus, and garlic. It is also frequently concentrated in commonly available high-fiber dietary supplements. Previous studies have found that inulin boosts populations of beneficial gut bacterial species which in turn boost levels of anti-inflammatory immune cells called regulatory T (Treg) cells.
In this new study, the researchers examined inulin’s effects more comprehensively. They gave mice an inulin-based, high-fiber diet for two weeks, and then analyzed the many differences between these mice and mice that had been fed a diet lacking inulin. A major difference was that the inulin diet, while increasing Treg cells, also induced markedly higher levels of white blood cells called eosinophils in the gut and lungs. A high level of eosinophils is a classic sign of type 2 inflammation and is typically seen in the setting of seasonal allergies and asthma.
Ultimately the researchers found that the eosinophil response was mediated by immune cells called group 2 innate lymphoid cells (ILC2s), which were activated by elevated levels of small molecules called bile acids in the blood. The bile acid levels were elevated due to the inulin-induced growth of certain bacterial species — a group called Bacteroidetes, found in both mice and humans — which have a bile acid-metabolizing enzyme.
“We were amazed to find such a strong association between inulin supplementation and increased bile acid levels,” Dr. Schroeder said. “We then found that deletion of the bile acid receptor abrogates the inulin-induced inflammation, suggesting that microbiota-driven changes in bile acid metabolism underlie the effects of inulin.”
“When we colonized germ-free mice (mice without microbiota) with one of these bacterial species, and then knocked out the gene for one bacterial enzyme that promotes bile acid production, the whole pathway leading from inulin to eosinophilia and allergic inflammation was blocked,” Dr. Guo said.
The finding that inulin promotes type 2 inflammation does not mean that this type of fiber is always “bad,” the researchers said. They found that inulin did worsen allergen-induced type 2 airway inflammation in mice. But the experiments also confirmed inulin’s previously reported effect at boosting anti-inflammatory Treg cells, which may in many cases, outweigh some pro-inflammatory impact. Moreover, a type 2 immune response, which in the gut and lungs involves an increased production of tissue-protecting mucus, is not necessarily harmful in healthy people — indeed, the researchers found in their mouse experiments that the inulin-induced type 2 inflammation enhances the defense against helminth infection.
“It could be that this inulin to type-2-inflammation pathway represents an adaptive, beneficial response to endemic helminth parasite infection, though its effects in a more industrialized, helminth-free environment are more complex and harder to predict,” said Dr. Arifuzzaman.
The researchers now plan to use their multi-disciplinary, multi-platform approach to study systematically the immune effects of the different types of dietary fiber as well as a range of other dietary supplements in different states of health and disease.
This work was supported in part by the National Institutes of Health (5T32HL134629, DP2 HD101401-01, AI140724, KL2 TR002385, R35 GM131877, DK126871, AI151599, AI095466, AI095608, AR070116, AI172027 and DK132244) the AGA Research Foundation, the WCM-RAPP Initiative, the W. M. Keck Foundation, the Howard Hughes Medical Institute, the LEO Foundation, CURE for IBD, the Jill Roberts Institute for Research in IBD, the Sanders Family Foundation and the Rosanne H. Silbermann Foundation.
Journal Reference:
- Mohammad Arifuzzaman, Tae Hyung Won, Ting-Ting Li, Hiroshi Yano, Sreehaas Digumarthi, Andrea F. Heras, Wen Zhang, Christopher N. Parkhurst, Sanchita Kashyap, Wen-Bing Jin, Gregory Garbès Putzel, Amy M. Tsou, Coco Chu, Qianru Wei, Alex Grier, Randy Longman, Gregory Sonnenberg, Ellen Scherl, Robbyn Sockolow, Dana Lukin, Robert Battat, Thomas Ciecierega, Aliza Solomon, Elaine Barfield, Kimberley Chien, Johanna Ferreira, Jasmin Williams, Shaira Khan, Peik Sean Chong, Samah Mozumder, Lance Chou, Wenqing Zhou, Anees Ahmed, Connie Zhong, Ann Joseph, Joseph Gladstone, Samantha Jensen, Stefan Worgall, Chun-Jun Guo, Frank C. Schroeder, David Artis. Inulin fibre promotes microbiota-derived bile acids and type 2 inflammation. Nature, 2022; DOI: 10.1038/s41586-022-05380-y
The study, published Nov. 2 in Nature, found that dietary inulin fiber alters the metabolism of certain gut bacteria, which in turn triggers what scientists call type 2 inflammation in the gut and lungs. This type of inflammation is thought to have evolved in mammals chiefly to defend against parasitic worm (“helminth”) infections, and is also part of normal wound-healing, although its inappropriate activation underlies allergies, asthma and other inflammatory diseases.
“There’s a lot to think about here, but, in general, these findings broaden our understanding of the relationship between diet, immunity, and the normally beneficial microorganisms that constitute our microbiota and colonize our bodies,” said study co-senior author Dr. David Artis, director of the Friedman Center for Nutrition and Inflammation and the Michael Kors Professor of Immunology at Weill Cornell Medicine.
The study’s scientific participants reflect the Friedman Center’s highly cross-collaborative research mission, drawing on expertise in bacterial genetics, biochemistry and immunology at Weill Cornell Medicine in New York City and Cornell’s Ithaca campus. Dr. Chun-Jun Guo, assistant professor of immunology in medicine at Weill Cornell Medicine, and Dr. Frank Schroeder, professor at the Boyce Thompson Institute and in the Department of Chemistry and Chemical Biology in the College of Arts and Sciences on Cornell’s Ithaca campus teamed up with the Artis laboratory to gain a detailed understanding of how an important dietary component affects the microbiome and the immune response. The study’s first author is Dr. Mohammad Arifuzzaman, a postdoctoral researcher in the Artis laboratory.Dr. Artis is also director of the Jill Roberts Institute for Inflammatory Bowel Disease at Weill Cornell Medicine.
Small amounts of inulin are present in a wide variety of fruits and vegetables, including bananas, asparagus, and garlic. It is also frequently concentrated in commonly available high-fiber dietary supplements. Previous studies have found that inulin boosts populations of beneficial gut bacterial species which in turn boost levels of anti-inflammatory immune cells called regulatory T (Treg) cells.
In this new study, the researchers examined inulin’s effects more comprehensively. They gave mice an inulin-based, high-fiber diet for two weeks, and then analyzed the many differences between these mice and mice that had been fed a diet lacking inulin. A major difference was that the inulin diet, while increasing Treg cells, also induced markedly higher levels of white blood cells called eosinophils in the gut and lungs. A high level of eosinophils is a classic sign of type 2 inflammation and is typically seen in the setting of seasonal allergies and asthma.
Ultimately the researchers found that the eosinophil response was mediated by immune cells called group 2 innate lymphoid cells (ILC2s), which were activated by elevated levels of small molecules called bile acids in the blood. The bile acid levels were elevated due to the inulin-induced growth of certain bacterial species — a group called Bacteroidetes, found in both mice and humans — which have a bile acid-metabolizing enzyme.
“We were amazed to find such a strong association between inulin supplementation and increased bile acid levels,” Dr. Schroeder said. “We then found that deletion of the bile acid receptor abrogates the inulin-induced inflammation, suggesting that microbiota-driven changes in bile acid metabolism underlie the effects of inulin.”
“When we colonized germ-free mice (mice without microbiota) with one of these bacterial species, and then knocked out the gene for one bacterial enzyme that promotes bile acid production, the whole pathway leading from inulin to eosinophilia and allergic inflammation was blocked,” Dr. Guo said.
The finding that inulin promotes type 2 inflammation does not mean that this type of fiber is always “bad,” the researchers said. They found that inulin did worsen allergen-induced type 2 airway inflammation in mice. But the experiments also confirmed inulin’s previously reported effect at boosting anti-inflammatory Treg cells, which may in many cases, outweigh some pro-inflammatory impact. Moreover, a type 2 immune response, which in the gut and lungs involves an increased production of tissue-protecting mucus, is not necessarily harmful in healthy people — indeed, the researchers found in their mouse experiments that the inulin-induced type 2 inflammation enhances the defense against helminth infection.
“It could be that this inulin to type-2-inflammation pathway represents an adaptive, beneficial response to endemic helminth parasite infection, though its effects in a more industrialized, helminth-free environment are more complex and harder to predict,” said Dr. Arifuzzaman.
The researchers now plan to use their multi-disciplinary, multi-platform approach to study systematically the immune effects of the different types of dietary fiber as well as a range of other dietary supplements in different states of health and disease.
This work was supported in part by the National Institutes of Health (5T32HL134629, DP2 HD101401-01, AI140724, KL2 TR002385, R35 GM131877, DK126871, AI151599, AI095466, AI095608, AR070116, AI172027 and DK132244) the AGA Research Foundation, the WCM-RAPP Initiative, the W. M. Keck Foundation, the Howard Hughes Medical Institute, the LEO Foundation, CURE for IBD, the Jill Roberts Institute for Research in IBD, the Sanders Family Foundation and the Rosanne H. Silbermann Foundation.
Source by
The post A common dietary fiber promotes allergy first appeared on .
)
[summary] => Journal Reference: Mohammad Arifuzzaman, Tae Hyung Won, Ting-Ting Li, Hiroshi Yano, Sreehaas Digumarthi, Andrea F. Heras, Wen Zhang, Christopher N. Parkhurst, Sanchita Kashyap, Wen-Bing Jin, Gregory Garbès Putzel, Amy M. Tsou, Coco Chu, Qianru Wei, Alex Grier, Randy Longman, Gregory Sonnenberg, Ellen Scherl, Robbyn Sockolow, Dana Lukin, Robert Battat, Thomas Ciecierega, Aliza Solomon, Elaine Barfield, ... Read more
The post A common dietary fiber promotes allergy first appeared on .
[atom_content] =>
Journal Reference:
- Mohammad Arifuzzaman, Tae Hyung Won, Ting-Ting Li, Hiroshi Yano, Sreehaas Digumarthi, Andrea F. Heras, Wen Zhang, Christopher N. Parkhurst, Sanchita Kashyap, Wen-Bing Jin, Gregory Garbès Putzel, Amy M. Tsou, Coco Chu, Qianru Wei, Alex Grier, Randy Longman, Gregory Sonnenberg, Ellen Scherl, Robbyn Sockolow, Dana Lukin, Robert Battat, Thomas Ciecierega, Aliza Solomon, Elaine Barfield, Kimberley Chien, Johanna Ferreira, Jasmin Williams, Shaira Khan, Peik Sean Chong, Samah Mozumder, Lance Chou, Wenqing Zhou, Anees Ahmed, Connie Zhong, Ann Joseph, Joseph Gladstone, Samantha Jensen, Stefan Worgall, Chun-Jun Guo, Frank C. Schroeder, David Artis. Inulin fibre promotes microbiota-derived bile acids and type 2 inflammation. Nature, 2022; DOI: 10.1038/s41586-022-05380-y
The study, published Nov. 2 in Nature, found that dietary inulin fiber alters the metabolism of certain gut bacteria, which in turn triggers what scientists call type 2 inflammation in the gut and lungs. This type of inflammation is thought to have evolved in mammals chiefly to defend against parasitic worm (“helminth”) infections, and is also part of normal wound-healing, although its inappropriate activation underlies allergies, asthma and other inflammatory diseases.
“There’s a lot to think about here, but, in general, these findings broaden our understanding of the relationship between diet, immunity, and the normally beneficial microorganisms that constitute our microbiota and colonize our bodies,” said study co-senior author Dr. David Artis, director of the Friedman Center for Nutrition and Inflammation and the Michael Kors Professor of Immunology at Weill Cornell Medicine.
The study’s scientific participants reflect the Friedman Center’s highly cross-collaborative research mission, drawing on expertise in bacterial genetics, biochemistry and immunology at Weill Cornell Medicine in New York City and Cornell’s Ithaca campus. Dr. Chun-Jun Guo, assistant professor of immunology in medicine at Weill Cornell Medicine, and Dr. Frank Schroeder, professor at the Boyce Thompson Institute and in the Department of Chemistry and Chemical Biology in the College of Arts and Sciences on Cornell’s Ithaca campus teamed up with the Artis laboratory to gain a detailed understanding of how an important dietary component affects the microbiome and the immune response. The study’s first author is Dr. Mohammad Arifuzzaman, a postdoctoral researcher in the Artis laboratory.Dr. Artis is also director of the Jill Roberts Institute for Inflammatory Bowel Disease at Weill Cornell Medicine.
Small amounts of inulin are present in a wide variety of fruits and vegetables, including bananas, asparagus, and garlic. It is also frequently concentrated in commonly available high-fiber dietary supplements. Previous studies have found that inulin boosts populations of beneficial gut bacterial species which in turn boost levels of anti-inflammatory immune cells called regulatory T (Treg) cells.
In this new study, the researchers examined inulin’s effects more comprehensively. They gave mice an inulin-based, high-fiber diet for two weeks, and then analyzed the many differences between these mice and mice that had been fed a diet lacking inulin. A major difference was that the inulin diet, while increasing Treg cells, also induced markedly higher levels of white blood cells called eosinophils in the gut and lungs. A high level of eosinophils is a classic sign of type 2 inflammation and is typically seen in the setting of seasonal allergies and asthma.
Ultimately the researchers found that the eosinophil response was mediated by immune cells called group 2 innate lymphoid cells (ILC2s), which were activated by elevated levels of small molecules called bile acids in the blood. The bile acid levels were elevated due to the inulin-induced growth of certain bacterial species — a group called Bacteroidetes, found in both mice and humans — which have a bile acid-metabolizing enzyme.
“We were amazed to find such a strong association between inulin supplementation and increased bile acid levels,” Dr. Schroeder said. “We then found that deletion of the bile acid receptor abrogates the inulin-induced inflammation, suggesting that microbiota-driven changes in bile acid metabolism underlie the effects of inulin.”
“When we colonized germ-free mice (mice without microbiota) with one of these bacterial species, and then knocked out the gene for one bacterial enzyme that promotes bile acid production, the whole pathway leading from inulin to eosinophilia and allergic inflammation was blocked,” Dr. Guo said.
The finding that inulin promotes type 2 inflammation does not mean that this type of fiber is always “bad,” the researchers said. They found that inulin did worsen allergen-induced type 2 airway inflammation in mice. But the experiments also confirmed inulin’s previously reported effect at boosting anti-inflammatory Treg cells, which may in many cases, outweigh some pro-inflammatory impact. Moreover, a type 2 immune response, which in the gut and lungs involves an increased production of tissue-protecting mucus, is not necessarily harmful in healthy people — indeed, the researchers found in their mouse experiments that the inulin-induced type 2 inflammation enhances the defense against helminth infection.
“It could be that this inulin to type-2-inflammation pathway represents an adaptive, beneficial response to endemic helminth parasite infection, though its effects in a more industrialized, helminth-free environment are more complex and harder to predict,” said Dr. Arifuzzaman.
The researchers now plan to use their multi-disciplinary, multi-platform approach to study systematically the immune effects of the different types of dietary fiber as well as a range of other dietary supplements in different states of health and disease.
This work was supported in part by the National Institutes of Health (5T32HL134629, DP2 HD101401-01, AI140724, KL2 TR002385, R35 GM131877, DK126871, AI151599, AI095466, AI095608, AR070116, AI172027 and DK132244) the AGA Research Foundation, the WCM-RAPP Initiative, the W. M. Keck Foundation, the Howard Hughes Medical Institute, the LEO Foundation, CURE for IBD, the Jill Roberts Institute for Research in IBD, the Sanders Family Foundation and the Rosanne H. Silbermann Foundation.
Journal Reference:
- Mohammad Arifuzzaman, Tae Hyung Won, Ting-Ting Li, Hiroshi Yano, Sreehaas Digumarthi, Andrea F. Heras, Wen Zhang, Christopher N. Parkhurst, Sanchita Kashyap, Wen-Bing Jin, Gregory Garbès Putzel, Amy M. Tsou, Coco Chu, Qianru Wei, Alex Grier, Randy Longman, Gregory Sonnenberg, Ellen Scherl, Robbyn Sockolow, Dana Lukin, Robert Battat, Thomas Ciecierega, Aliza Solomon, Elaine Barfield, Kimberley Chien, Johanna Ferreira, Jasmin Williams, Shaira Khan, Peik Sean Chong, Samah Mozumder, Lance Chou, Wenqing Zhou, Anees Ahmed, Connie Zhong, Ann Joseph, Joseph Gladstone, Samantha Jensen, Stefan Worgall, Chun-Jun Guo, Frank C. Schroeder, David Artis. Inulin fibre promotes microbiota-derived bile acids and type 2 inflammation. Nature, 2022; DOI: 10.1038/s41586-022-05380-y
The study, published Nov. 2 in Nature, found that dietary inulin fiber alters the metabolism of certain gut bacteria, which in turn triggers what scientists call type 2 inflammation in the gut and lungs. This type of inflammation is thought to have evolved in mammals chiefly to defend against parasitic worm (“helminth”) infections, and is also part of normal wound-healing, although its inappropriate activation underlies allergies, asthma and other inflammatory diseases.
“There’s a lot to think about here, but, in general, these findings broaden our understanding of the relationship between diet, immunity, and the normally beneficial microorganisms that constitute our microbiota and colonize our bodies,” said study co-senior author Dr. David Artis, director of the Friedman Center for Nutrition and Inflammation and the Michael Kors Professor of Immunology at Weill Cornell Medicine.
The study’s scientific participants reflect the Friedman Center’s highly cross-collaborative research mission, drawing on expertise in bacterial genetics, biochemistry and immunology at Weill Cornell Medicine in New York City and Cornell’s Ithaca campus. Dr. Chun-Jun Guo, assistant professor of immunology in medicine at Weill Cornell Medicine, and Dr. Frank Schroeder, professor at the Boyce Thompson Institute and in the Department of Chemistry and Chemical Biology in the College of Arts and Sciences on Cornell’s Ithaca campus teamed up with the Artis laboratory to gain a detailed understanding of how an important dietary component affects the microbiome and the immune response. The study’s first author is Dr. Mohammad Arifuzzaman, a postdoctoral researcher in the Artis laboratory.Dr. Artis is also director of the Jill Roberts Institute for Inflammatory Bowel Disease at Weill Cornell Medicine.
Small amounts of inulin are present in a wide variety of fruits and vegetables, including bananas, asparagus, and garlic. It is also frequently concentrated in commonly available high-fiber dietary supplements. Previous studies have found that inulin boosts populations of beneficial gut bacterial species which in turn boost levels of anti-inflammatory immune cells called regulatory T (Treg) cells.
In this new study, the researchers examined inulin’s effects more comprehensively. They gave mice an inulin-based, high-fiber diet for two weeks, and then analyzed the many differences between these mice and mice that had been fed a diet lacking inulin. A major difference was that the inulin diet, while increasing Treg cells, also induced markedly higher levels of white blood cells called eosinophils in the gut and lungs. A high level of eosinophils is a classic sign of type 2 inflammation and is typically seen in the setting of seasonal allergies and asthma.
Ultimately the researchers found that the eosinophil response was mediated by immune cells called group 2 innate lymphoid cells (ILC2s), which were activated by elevated levels of small molecules called bile acids in the blood. The bile acid levels were elevated due to the inulin-induced growth of certain bacterial species — a group called Bacteroidetes, found in both mice and humans — which have a bile acid-metabolizing enzyme.
“We were amazed to find such a strong association between inulin supplementation and increased bile acid levels,” Dr. Schroeder said. “We then found that deletion of the bile acid receptor abrogates the inulin-induced inflammation, suggesting that microbiota-driven changes in bile acid metabolism underlie the effects of inulin.”
“When we colonized germ-free mice (mice without microbiota) with one of these bacterial species, and then knocked out the gene for one bacterial enzyme that promotes bile acid production, the whole pathway leading from inulin to eosinophilia and allergic inflammation was blocked,” Dr. Guo said.
The finding that inulin promotes type 2 inflammation does not mean that this type of fiber is always “bad,” the researchers said. They found that inulin did worsen allergen-induced type 2 airway inflammation in mice. But the experiments also confirmed inulin’s previously reported effect at boosting anti-inflammatory Treg cells, which may in many cases, outweigh some pro-inflammatory impact. Moreover, a type 2 immune response, which in the gut and lungs involves an increased production of tissue-protecting mucus, is not necessarily harmful in healthy people — indeed, the researchers found in their mouse experiments that the inulin-induced type 2 inflammation enhances the defense against helminth infection.
“It could be that this inulin to type-2-inflammation pathway represents an adaptive, beneficial response to endemic helminth parasite infection, though its effects in a more industrialized, helminth-free environment are more complex and harder to predict,” said Dr. Arifuzzaman.
The researchers now plan to use their multi-disciplinary, multi-platform approach to study systematically the immune effects of the different types of dietary fiber as well as a range of other dietary supplements in different states of health and disease.
This work was supported in part by the National Institutes of Health (5T32HL134629, DP2 HD101401-01, AI140724, KL2 TR002385, R35 GM131877, DK126871, AI151599, AI095466, AI095608, AR070116, AI172027 and DK132244) the AGA Research Foundation, the WCM-RAPP Initiative, the W. M. Keck Foundation, the Howard Hughes Medical Institute, the LEO Foundation, CURE for IBD, the Jill Roberts Institute for Research in IBD, the Sanders Family Foundation and the Rosanne H. Silbermann Foundation.
Source by
The post A common dietary fiber promotes allergy first appeared on .
)
[8] => Array
(
[title] => Looking for romance? That first impression matters
[link] => https://coolnspicy.com/science/looking-for-romance-that-first-impression-matters/
[dc] => Array
(
[creator] => Michael Steiner
)
[pubdate] => Thu, 03 Nov 2022 08:00:19 +0000
[category] => scienceimpressionmattersromance
[guid] => https://coolnspicy.com/?p=281
[description] => Journal Reference: Alexander Baxter, Jessica A. Maxwell, Karen L. Bales, Eli J. Finkel, Emily A. Impett, Paul W. Eastwick. Initial impressions of compatibility and mate value predict later dating and romantic interest. Proceedings of the National Academy of Sciences, 2022; 119 (45) DOI: 10.1073/pnas.2206925119 Although popularity and compatibility have been studied in established romantic relationships, ... Read more
The post Looking for romance? That first impression matters first appeared on .
[content] => Array
(
[encoded] =>
Journal Reference:
- Alexander Baxter, Jessica A. Maxwell, Karen L. Bales, Eli J. Finkel, Emily A. Impett, Paul W. Eastwick. Initial impressions of compatibility and mate value predict later dating and romantic interest. Proceedings of the National Academy of Sciences, 2022; 119 (45) DOI: 10.1073/pnas.2206925119
Although popularity and compatibility have been studied in established romantic relationships, in one of the first studies of its kind, UC Davis researchers explored whether these and other types of romantic first impressions affected later romantic outcomes. Researchers found that first impressions tend to linger, shaping whether people desired further contact with potential romantic partners after an initial meeting.
The study was published Oct. 31in the Proceedings of the National Academy of Sciences.
“Although we expected popularity to be an important factor in the study, we were amazed to find that a good first impression is not just a popularity contest, it’s also about compatibility, even when people are still getting to know each other,” said Alexander Baxter, a UC Davis doctoral student in psychology and co-author of the study. “In other words, although it helps to be popular when it comes to getting a second date, having a unique connection with a potential partner can be just as important.”
Assessed during speed dates
The researchers asked more than 550 speed-daters, including some men who date men, to rate their romantic interest in the potential partners they met. The participants, all from the United States or Canada, included both college students and people attending a comic book convention, who cumulatively attended more than 6,600 speed-dates during the experiment.
“Representation really matters in psychology research, and one of the strengths of our study is that we included a subsample of men who date men that attended an all-male speed-dating event,” Baxter said. “This means that our findings generalize not only to male-female relationships, but also to male-male relationships too. We hope that future studies will consider other diverse types of relationships.”
After the speed-dating event, researchers surveyed the participants over the next two to three months to assess whether they dated any of the potential partners that they met and how their romantic feelings changed over time.
Researchers used a statistical model to test whether later romantic outcomes were predicted by three factors that affect how romantic first impressions form — selectivity, popularity and compatibility. In other words, they looked at patterns of initial desire that were observed during the speed-dates, and assessed whether these factors differently predicted whether people later pursued a relationship with the potential partners that they met.
The factors assessed were:
- Selectivity — Daniel liked Rose because he liked everyone;
- Popularity — Daniel liked Rose because everyone liked her; and
- Compatibility — Daniel uniquely liked Rose, above and beyond his own flirty disposition and her general popularity.
The results showed that people were particularly likely to pursue a romantic relationship with those who were popular and those they were compatible with. Selectivity played a relatively small role, with more romantically outgoing individuals being slightly more likely than less outgoing people to pursue their speed-dating matches, according to the study.
In addition to Baxter, who is also a researcher with the California National Primate Research Center at UC Davis, co-authors of the study include Paul Eastwick, professor of psychology and director of the Attraction and Relationships Research Laboratory, UC Davis; Professor Karen L. Bales, of the primate center and Department of Neurobiology, Physiology, and Behavior, and Department of Psychology, UC Davis; Jessica A. Maxwell, School of Psychology, University of Auckland, Auckland, New Zealand; Eli J. Finkel, Department of Psychology, Kellogg School of Management, Northwestern University, Evanston, Illinois; and Emily A. Impett, Department of Psychology, University of Toronto Mississauga, Ontario, Canada.
This work was supported by a grant from The Love Consortium, The John Templeton Foundation (grant 61280) and the National Science Foundation.
Journal Reference:
- Alexander Baxter, Jessica A. Maxwell, Karen L. Bales, Eli J. Finkel, Emily A. Impett, Paul W. Eastwick. Initial impressions of compatibility and mate value predict later dating and romantic interest. Proceedings of the National Academy of Sciences, 2022; 119 (45) DOI: 10.1073/pnas.2206925119
Although popularity and compatibility have been studied in established romantic relationships, in one of the first studies of its kind, UC Davis researchers explored whether these and other types of romantic first impressions affected later romantic outcomes. Researchers found that first impressions tend to linger, shaping whether people desired further contact with potential romantic partners after an initial meeting.
The study was published Oct. 31in the Proceedings of the National Academy of Sciences.
“Although we expected popularity to be an important factor in the study, we were amazed to find that a good first impression is not just a popularity contest, it’s also about compatibility, even when people are still getting to know each other,” said Alexander Baxter, a UC Davis doctoral student in psychology and co-author of the study. “In other words, although it helps to be popular when it comes to getting a second date, having a unique connection with a potential partner can be just as important.”
Assessed during speed dates
The researchers asked more than 550 speed-daters, including some men who date men, to rate their romantic interest in the potential partners they met. The participants, all from the United States or Canada, included both college students and people attending a comic book convention, who cumulatively attended more than 6,600 speed-dates during the experiment.
“Representation really matters in psychology research, and one of the strengths of our study is that we included a subsample of men who date men that attended an all-male speed-dating event,” Baxter said. “This means that our findings generalize not only to male-female relationships, but also to male-male relationships too. We hope that future studies will consider other diverse types of relationships.”
After the speed-dating event, researchers surveyed the participants over the next two to three months to assess whether they dated any of the potential partners that they met and how their romantic feelings changed over time.
Researchers used a statistical model to test whether later romantic outcomes were predicted by three factors that affect how romantic first impressions form — selectivity, popularity and compatibility. In other words, they looked at patterns of initial desire that were observed during the speed-dates, and assessed whether these factors differently predicted whether people later pursued a relationship with the potential partners that they met.
The factors assessed were:
- Selectivity — Daniel liked Rose because he liked everyone;
- Popularity — Daniel liked Rose because everyone liked her; and
- Compatibility — Daniel uniquely liked Rose, above and beyond his own flirty disposition and her general popularity.
The results showed that people were particularly likely to pursue a romantic relationship with those who were popular and those they were compatible with. Selectivity played a relatively small role, with more romantically outgoing individuals being slightly more likely than less outgoing people to pursue their speed-dating matches, according to the study.
In addition to Baxter, who is also a researcher with the California National Primate Research Center at UC Davis, co-authors of the study include Paul Eastwick, professor of psychology and director of the Attraction and Relationships Research Laboratory, UC Davis; Professor Karen L. Bales, of the primate center and Department of Neurobiology, Physiology, and Behavior, and Department of Psychology, UC Davis; Jessica A. Maxwell, School of Psychology, University of Auckland, Auckland, New Zealand; Eli J. Finkel, Department of Psychology, Kellogg School of Management, Northwestern University, Evanston, Illinois; and Emily A. Impett, Department of Psychology, University of Toronto Mississauga, Ontario, Canada.
This work was supported by a grant from The Love Consortium, The John Templeton Foundation (grant 61280) and the National Science Foundation.
Looking for romance? That first impression matters first appeared on .
)
[summary] => Journal Reference: Alexander Baxter, Jessica A. Maxwell, Karen L. Bales, Eli J. Finkel, Emily A. Impett, Paul W. Eastwick. Initial impressions of compatibility and mate value predict later dating and romantic interest. Proceedings of the National Academy of Sciences, 2022; 119 (45) DOI: 10.1073/pnas.2206925119 Although popularity and compatibility have been studied in established romantic relationships, ... Read more
The post Looking for romance? That first impression matters first appeared on .
[atom_content] =>
Journal Reference:
- Alexander Baxter, Jessica A. Maxwell, Karen L. Bales, Eli J. Finkel, Emily A. Impett, Paul W. Eastwick. Initial impressions of compatibility and mate value predict later dating and romantic interest. Proceedings of the National Academy of Sciences, 2022; 119 (45) DOI: 10.1073/pnas.2206925119
Although popularity and compatibility have been studied in established romantic relationships, in one of the first studies of its kind, UC Davis researchers explored whether these and other types of romantic first impressions affected later romantic outcomes. Researchers found that first impressions tend to linger, shaping whether people desired further contact with potential romantic partners after an initial meeting.
The study was published Oct. 31in the Proceedings of the National Academy of Sciences.
“Although we expected popularity to be an important factor in the study, we were amazed to find that a good first impression is not just a popularity contest, it’s also about compatibility, even when people are still getting to know each other,” said Alexander Baxter, a UC Davis doctoral student in psychology and co-author of the study. “In other words, although it helps to be popular when it comes to getting a second date, having a unique connection with a potential partner can be just as important.”
Assessed during speed dates
The researchers asked more than 550 speed-daters, including some men who date men, to rate their romantic interest in the potential partners they met. The participants, all from the United States or Canada, included both college students and people attending a comic book convention, who cumulatively attended more than 6,600 speed-dates during the experiment.
“Representation really matters in psychology research, and one of the strengths of our study is that we included a subsample of men who date men that attended an all-male speed-dating event,” Baxter said. “This means that our findings generalize not only to male-female relationships, but also to male-male relationships too. We hope that future studies will consider other diverse types of relationships.”
After the speed-dating event, researchers surveyed the participants over the next two to three months to assess whether they dated any of the potential partners that they met and how their romantic feelings changed over time.
Researchers used a statistical model to test whether later romantic outcomes were predicted by three factors that affect how romantic first impressions form — selectivity, popularity and compatibility. In other words, they looked at patterns of initial desire that were observed during the speed-dates, and assessed whether these factors differently predicted whether people later pursued a relationship with the potential partners that they met.
The factors assessed were:
- Selectivity — Daniel liked Rose because he liked everyone;
- Popularity — Daniel liked Rose because everyone liked her; and
- Compatibility — Daniel uniquely liked Rose, above and beyond his own flirty disposition and her general popularity.
The results showed that people were particularly likely to pursue a romantic relationship with those who were popular and those they were compatible with. Selectivity played a relatively small role, with more romantically outgoing individuals being slightly more likely than less outgoing people to pursue their speed-dating matches, according to the study.
In addition to Baxter, who is also a researcher with the California National Primate Research Center at UC Davis, co-authors of the study include Paul Eastwick, professor of psychology and director of the Attraction and Relationships Research Laboratory, UC Davis; Professor Karen L. Bales, of the primate center and Department of Neurobiology, Physiology, and Behavior, and Department of Psychology, UC Davis; Jessica A. Maxwell, School of Psychology, University of Auckland, Auckland, New Zealand; Eli J. Finkel, Department of Psychology, Kellogg School of Management, Northwestern University, Evanston, Illinois; and Emily A. Impett, Department of Psychology, University of Toronto Mississauga, Ontario, Canada.
This work was supported by a grant from The Love Consortium, The John Templeton Foundation (grant 61280) and the National Science Foundation.
Journal Reference:
- Alexander Baxter, Jessica A. Maxwell, Karen L. Bales, Eli J. Finkel, Emily A. Impett, Paul W. Eastwick. Initial impressions of compatibility and mate value predict later dating and romantic interest. Proceedings of the National Academy of Sciences, 2022; 119 (45) DOI: 10.1073/pnas.2206925119
Although popularity and compatibility have been studied in established romantic relationships, in one of the first studies of its kind, UC Davis researchers explored whether these and other types of romantic first impressions affected later romantic outcomes. Researchers found that first impressions tend to linger, shaping whether people desired further contact with potential romantic partners after an initial meeting.
The study was published Oct. 31in the Proceedings of the National Academy of Sciences.
“Although we expected popularity to be an important factor in the study, we were amazed to find that a good first impression is not just a popularity contest, it’s also about compatibility, even when people are still getting to know each other,” said Alexander Baxter, a UC Davis doctoral student in psychology and co-author of the study. “In other words, although it helps to be popular when it comes to getting a second date, having a unique connection with a potential partner can be just as important.”
Assessed during speed dates
The researchers asked more than 550 speed-daters, including some men who date men, to rate their romantic interest in the potential partners they met. The participants, all from the United States or Canada, included both college students and people attending a comic book convention, who cumulatively attended more than 6,600 speed-dates during the experiment.
“Representation really matters in psychology research, and one of the strengths of our study is that we included a subsample of men who date men that attended an all-male speed-dating event,” Baxter said. “This means that our findings generalize not only to male-female relationships, but also to male-male relationships too. We hope that future studies will consider other diverse types of relationships.”
After the speed-dating event, researchers surveyed the participants over the next two to three months to assess whether they dated any of the potential partners that they met and how their romantic feelings changed over time.
Researchers used a statistical model to test whether later romantic outcomes were predicted by three factors that affect how romantic first impressions form — selectivity, popularity and compatibility. In other words, they looked at patterns of initial desire that were observed during the speed-dates, and assessed whether these factors differently predicted whether people later pursued a relationship with the potential partners that they met.
The factors assessed were:
- Selectivity — Daniel liked Rose because he liked everyone;
- Popularity — Daniel liked Rose because everyone liked her; and
- Compatibility — Daniel uniquely liked Rose, above and beyond his own flirty disposition and her general popularity.
The results showed that people were particularly likely to pursue a romantic relationship with those who were popular and those they were compatible with. Selectivity played a relatively small role, with more romantically outgoing individuals being slightly more likely than less outgoing people to pursue their speed-dating matches, according to the study.
In addition to Baxter, who is also a researcher with the California National Primate Research Center at UC Davis, co-authors of the study include Paul Eastwick, professor of psychology and director of the Attraction and Relationships Research Laboratory, UC Davis; Professor Karen L. Bales, of the primate center and Department of Neurobiology, Physiology, and Behavior, and Department of Psychology, UC Davis; Jessica A. Maxwell, School of Psychology, University of Auckland, Auckland, New Zealand; Eli J. Finkel, Department of Psychology, Kellogg School of Management, Northwestern University, Evanston, Illinois; and Emily A. Impett, Department of Psychology, University of Toronto Mississauga, Ontario, Canada.
This work was supported by a grant from The Love Consortium, The John Templeton Foundation (grant 61280) and the National Science Foundation.
Looking for romance? That first impression matters first appeared on .
)
[9] => Array
(
[title] => How network pruning can skew deep learning models
[link] => https://coolnspicy.com/science/how-network-pruning-can-skew-deep-learning-models/
[dc] => Array
(
[creator] => Michael Steiner
)
[pubdate] => Thu, 03 Nov 2022 07:58:46 +0000
[category] => sciencedeeplearningmodelsnetworkpruningskew
[guid] => https://coolnspicy.com/?p=279
[description] => Deep learning is a type of artificial intelligence that can be used to classify things, such as images, text or sound. For example, it can be used to identify individuals based on facial images. However, deep learning models often require a lot of computing resources to operate. This poses challenges when a deep learning model ... Read more
The post How network pruning can skew deep learning models first appeared on .
[content] => Array
(
[encoded] =>
Deep learning is a type of artificial intelligence that can be used to classify things, such as images, text or sound. For example, it can be used to identify individuals based on facial images. However, deep learning models often require a lot of computing resources to operate. This poses challenges when a deep learning model is put into practice for some applications.
To address these challenges, some systems engage in “neural network pruning.” This effectively makes the deep learning model more compact and, therefore, able to operate while using fewer computing resources.
“However, our research shows that this network pruning can impair the ability of deep learning models to identify some groups,” says Jung-Eun Kim, co-author of a paper on the work and an assistant professor of computer science at North Carolina State University.
“For example, if a security system uses deep learning to scan people’s faces in order to determine whether they have access to a building, the deep learning model would have to be made compact so that it can operate efficiently. This may work fine most of the time, but the network pruning could also affect the deep learning model’s ability to identify some faces.”
In their new paper, the researchers lay out why network pruning can adversely affect the performance of the model at identifying certain groups — which the literature calls “minority groups” — and demonstrate a new technique for addressing these challenges.
Two factors explain how network pruning can impair the performance of deep learning models.
In technical terms, these two factors are: disparity in gradient norms across groups; and disparity in Hessian norms associated with inaccuracies of a group’s data. In practical terms, this means that deep learning models can become less accurate in recognizing specific categories of images, sounds or text. Specifically, the network pruning can amplify accuracy deficiencies that already existed in the model.
For example, if a deep learning model is trained to recognize faces using a data set that includes the faces of 100 white people and 60 Asian people, it might be more accurate at recognizing white faces, but could still achieve adequate performance for recognizing Asian faces. After network pruning, the model is more likely to be unable to recognize some Asian faces.
“The deficiency may not have been noticeable in the original model, but because it’s amplified by the network pruning, the deficiency may become noticeable,” Kim says.
“To mitigate this problem, we’ve demonstrated an approach that uses mathematical techniques to equalize the groups that the deep learning model is using to categorize data samples,” Kim says. “In other words, we are using algorithms to address the gap in accuracy across groups.”
In testing, the researchers demonstrated that using their mitigation technique improved the fairness of a deep learning model that had undergone network pruning, essentially returning it to pre-pruning levels of accuracy.
“I think the most important aspect of this work is that we now have a more thorough understanding of exactly how network pruning can influence the performance of deep learning models to identify minority groups, both theoretically and empirically,” Kim says. “We’re also open to working with partners to identify unknown or overlooked impacts of model reduction techniques, particularly in real-world applications for deep learning models.”
The paper, “Pruning Has a Disparate Impact on Model Accuracy,” will be presented at the 36th Conference on Neural Information Processing Systems (NeurIPS 2022), being held Nov. 28-Dec. 9 in New Orleans. First author of the paper is Cuong Tran of Syracuse University. The paper was co-authored by Ferdinando Fioretto of Syracuse, and by Rakshit Naidu of Carnegie Mellon University.
The work was done with support from the National Science Foundation, under grants SaTC-1945541, SaTC-2133169 and CAREER-2143706; as well as a Google Research Scholar Award and an Amazon Research Award.
Deep learning is a type of artificial intelligence that can be used to classify things, such as images, text or sound. For example, it can be used to identify individuals based on facial images. However, deep learning models often require a lot of computing resources to operate. This poses challenges when a deep learning model is put into practice for some applications.
To address these challenges, some systems engage in “neural network pruning.” This effectively makes the deep learning model more compact and, therefore, able to operate while using fewer computing resources.
“However, our research shows that this network pruning can impair the ability of deep learning models to identify some groups,” says Jung-Eun Kim, co-author of a paper on the work and an assistant professor of computer science at North Carolina State University.
“For example, if a security system uses deep learning to scan people’s faces in order to determine whether they have access to a building, the deep learning model would have to be made compact so that it can operate efficiently. This may work fine most of the time, but the network pruning could also affect the deep learning model’s ability to identify some faces.”
In their new paper, the researchers lay out why network pruning can adversely affect the performance of the model at identifying certain groups — which the literature calls “minority groups” — and demonstrate a new technique for addressing these challenges.
Two factors explain how network pruning can impair the performance of deep learning models.
In technical terms, these two factors are: disparity in gradient norms across groups; and disparity in Hessian norms associated with inaccuracies of a group’s data. In practical terms, this means that deep learning models can become less accurate in recognizing specific categories of images, sounds or text. Specifically, the network pruning can amplify accuracy deficiencies that already existed in the model.
For example, if a deep learning model is trained to recognize faces using a data set that includes the faces of 100 white people and 60 Asian people, it might be more accurate at recognizing white faces, but could still achieve adequate performance for recognizing Asian faces. After network pruning, the model is more likely to be unable to recognize some Asian faces.
“The deficiency may not have been noticeable in the original model, but because it’s amplified by the network pruning, the deficiency may become noticeable,” Kim says.
“To mitigate this problem, we’ve demonstrated an approach that uses mathematical techniques to equalize the groups that the deep learning model is using to categorize data samples,” Kim says. “In other words, we are using algorithms to address the gap in accuracy across groups.”
In testing, the researchers demonstrated that using their mitigation technique improved the fairness of a deep learning model that had undergone network pruning, essentially returning it to pre-pruning levels of accuracy.
“I think the most important aspect of this work is that we now have a more thorough understanding of exactly how network pruning can influence the performance of deep learning models to identify minority groups, both theoretically and empirically,” Kim says. “We’re also open to working with partners to identify unknown or overlooked impacts of model reduction techniques, particularly in real-world applications for deep learning models.”
The paper, “Pruning Has a Disparate Impact on Model Accuracy,” will be presented at the 36th Conference on Neural Information Processing Systems (NeurIPS 2022), being held Nov. 28-Dec. 9 in New Orleans. First author of the paper is Cuong Tran of Syracuse University. The paper was co-authored by Ferdinando Fioretto of Syracuse, and by Rakshit Naidu of Carnegie Mellon University.
The work was done with support from the National Science Foundation, under grants SaTC-1945541, SaTC-2133169 and CAREER-2143706; as well as a Google Research Scholar Award and an Amazon Research Award.
Source link How network pruning can skew deep learning models
The post How network pruning can skew deep learning models first appeared on .
)
[summary] => Deep learning is a type of artificial intelligence that can be used to classify things, such as images, text or sound. For example, it can be used to identify individuals based on facial images. However, deep learning models often require a lot of computing resources to operate. This poses challenges when a deep learning model ... Read more
The post How network pruning can skew deep learning models first appeared on .
[atom_content] =>
Deep learning is a type of artificial intelligence that can be used to classify things, such as images, text or sound. For example, it can be used to identify individuals based on facial images. However, deep learning models often require a lot of computing resources to operate. This poses challenges when a deep learning model is put into practice for some applications.
To address these challenges, some systems engage in “neural network pruning.” This effectively makes the deep learning model more compact and, therefore, able to operate while using fewer computing resources.
“However, our research shows that this network pruning can impair the ability of deep learning models to identify some groups,” says Jung-Eun Kim, co-author of a paper on the work and an assistant professor of computer science at North Carolina State University.
“For example, if a security system uses deep learning to scan people’s faces in order to determine whether they have access to a building, the deep learning model would have to be made compact so that it can operate efficiently. This may work fine most of the time, but the network pruning could also affect the deep learning model’s ability to identify some faces.”
In their new paper, the researchers lay out why network pruning can adversely affect the performance of the model at identifying certain groups — which the literature calls “minority groups” — and demonstrate a new technique for addressing these challenges.
Two factors explain how network pruning can impair the performance of deep learning models.
In technical terms, these two factors are: disparity in gradient norms across groups; and disparity in Hessian norms associated with inaccuracies of a group’s data. In practical terms, this means that deep learning models can become less accurate in recognizing specific categories of images, sounds or text. Specifically, the network pruning can amplify accuracy deficiencies that already existed in the model.
For example, if a deep learning model is trained to recognize faces using a data set that includes the faces of 100 white people and 60 Asian people, it might be more accurate at recognizing white faces, but could still achieve adequate performance for recognizing Asian faces. After network pruning, the model is more likely to be unable to recognize some Asian faces.
“The deficiency may not have been noticeable in the original model, but because it’s amplified by the network pruning, the deficiency may become noticeable,” Kim says.
“To mitigate this problem, we’ve demonstrated an approach that uses mathematical techniques to equalize the groups that the deep learning model is using to categorize data samples,” Kim says. “In other words, we are using algorithms to address the gap in accuracy across groups.”
In testing, the researchers demonstrated that using their mitigation technique improved the fairness of a deep learning model that had undergone network pruning, essentially returning it to pre-pruning levels of accuracy.
“I think the most important aspect of this work is that we now have a more thorough understanding of exactly how network pruning can influence the performance of deep learning models to identify minority groups, both theoretically and empirically,” Kim says. “We’re also open to working with partners to identify unknown or overlooked impacts of model reduction techniques, particularly in real-world applications for deep learning models.”
The paper, “Pruning Has a Disparate Impact on Model Accuracy,” will be presented at the 36th Conference on Neural Information Processing Systems (NeurIPS 2022), being held Nov. 28-Dec. 9 in New Orleans. First author of the paper is Cuong Tran of Syracuse University. The paper was co-authored by Ferdinando Fioretto of Syracuse, and by Rakshit Naidu of Carnegie Mellon University.
The work was done with support from the National Science Foundation, under grants SaTC-1945541, SaTC-2133169 and CAREER-2143706; as well as a Google Research Scholar Award and an Amazon Research Award.
Deep learning is a type of artificial intelligence that can be used to classify things, such as images, text or sound. For example, it can be used to identify individuals based on facial images. However, deep learning models often require a lot of computing resources to operate. This poses challenges when a deep learning model is put into practice for some applications.
To address these challenges, some systems engage in “neural network pruning.” This effectively makes the deep learning model more compact and, therefore, able to operate while using fewer computing resources.
“However, our research shows that this network pruning can impair the ability of deep learning models to identify some groups,” says Jung-Eun Kim, co-author of a paper on the work and an assistant professor of computer science at North Carolina State University.
“For example, if a security system uses deep learning to scan people’s faces in order to determine whether they have access to a building, the deep learning model would have to be made compact so that it can operate efficiently. This may work fine most of the time, but the network pruning could also affect the deep learning model’s ability to identify some faces.”
In their new paper, the researchers lay out why network pruning can adversely affect the performance of the model at identifying certain groups — which the literature calls “minority groups” — and demonstrate a new technique for addressing these challenges.
Two factors explain how network pruning can impair the performance of deep learning models.
In technical terms, these two factors are: disparity in gradient norms across groups; and disparity in Hessian norms associated with inaccuracies of a group’s data. In practical terms, this means that deep learning models can become less accurate in recognizing specific categories of images, sounds or text. Specifically, the network pruning can amplify accuracy deficiencies that already existed in the model.
For example, if a deep learning model is trained to recognize faces using a data set that includes the faces of 100 white people and 60 Asian people, it might be more accurate at recognizing white faces, but could still achieve adequate performance for recognizing Asian faces. After network pruning, the model is more likely to be unable to recognize some Asian faces.
“The deficiency may not have been noticeable in the original model, but because it’s amplified by the network pruning, the deficiency may become noticeable,” Kim says.
“To mitigate this problem, we’ve demonstrated an approach that uses mathematical techniques to equalize the groups that the deep learning model is using to categorize data samples,” Kim says. “In other words, we are using algorithms to address the gap in accuracy across groups.”
In testing, the researchers demonstrated that using their mitigation technique improved the fairness of a deep learning model that had undergone network pruning, essentially returning it to pre-pruning levels of accuracy.
“I think the most important aspect of this work is that we now have a more thorough understanding of exactly how network pruning can influence the performance of deep learning models to identify minority groups, both theoretically and empirically,” Kim says. “We’re also open to working with partners to identify unknown or overlooked impacts of model reduction techniques, particularly in real-world applications for deep learning models.”
The paper, “Pruning Has a Disparate Impact on Model Accuracy,” will be presented at the 36th Conference on Neural Information Processing Systems (NeurIPS 2022), being held Nov. 28-Dec. 9 in New Orleans. First author of the paper is Cuong Tran of Syracuse University. The paper was co-authored by Ferdinando Fioretto of Syracuse, and by Rakshit Naidu of Carnegie Mellon University.
The work was done with support from the National Science Foundation, under grants SaTC-1945541, SaTC-2133169 and CAREER-2143706; as well as a Google Research Scholar Award and an Amazon Research Award.
Source link How network pruning can skew deep learning models
The post How network pruning can skew deep learning models first appeared on .
)
)
[channel] => Array
(
[link] => https://coolnspicy.com
[lastbuilddate] => Thu, 03 Nov 2022 08:17:48 +0000
[language] => en-US
[sy] => Array
(
[updateperiod] =>
hourly
[updatefrequency] =>
1
)
[generator] => https://wordpress.org/?v=6.1
[tagline] =>
)
[textinput] => Array
(
)
[image] => Array
(
)
[feed_type] => RSS
[feed_version] => 2.0
[stack] => Array
(
)
[inchannel] =>
[initem] =>
[incontent] =>
[intextinput] =>
[inimage] =>
[current_field] =>
[current_namespace] =>
[ERROR] =>
[_CONTENT_CONSTRUCTS] => Array
(
[0] => content
[1] => summary
[2] => info
[3] => title
[4] => tagline
[5] => copyright
)
[last_modified] => Thu, 03 Nov 2022 08:17:48 GMT
[etag] => "93f2f7611dfbaa6ec2b9536599bb82e1-gzip"
)