Underspecification Presents Challenges for Credibility in Modern Machine Learning (Paper Explained)
#ai #research #machinelearning
Deep Learning models are often overparameterized and have many degrees of freedom, which leads to many local minima that all perform equally well on the test set. But it turns out that even though they all generalize in-distribution, the performance of these models can be drastically different when tested out-of-distribution. Notably, in many cases, a good model can actually be found among all these candidates, but it seems impossible to select it. This paper describes this problem, which it calls underspecification, and gives several theoretical and practical examples.
OUTLINE:
0:00 - Into & Overview
2:00 - Underspecification of ML Pipelines
11:15 - Stress Tests
12:40 - Epidemiological Example
20:45 - Theoretical Model
26:55 - Example from Medical Genomics
34:00 - ImageNet-C Example
36:50 - BERT Models
56:55 - Conclusion & Comments
Paper: https://arxiv.org/abs/2011.03395
Abstract:
ML models often exhibit unexpectedly poor behavior when they are deployed in real-world domains. We identify underspecification as a key reason for these failures. An ML pipeline is underspecified when it can return many predictors with equivalently strong held-out performance in the training domain. Underspecification is common in modern ML pipelines, such as those based on deep learning. Predictors returned by underspecified pipelines are often treated as equivalent based on their training domain performance, but we show here that such predictors can behave very differently in deployment domains. This ambiguity can lead to instability and poor model behavior in practice, and is a distinct failure mode from previously identified issues arising from structural mismatch between training and deployment domains. We show that this problem appears in a wide variety of practical ML pipelines, using examples from computer vision, medical imaging, natural language processing, clinical risk prediction based on electronic health records, and medical genomics. Our results show the need to explicitly account for underspecification in modeling pipelines that are intended for real-world deployment in any domain.
Authors: Alexander D'Amour, Katherine Heller, Dan Moldovan, Ben Adlam, Babak Alipanahi, Alex Beutel, Christina Chen, Jonathan Deaton, Jacob Eisenstein, Matthew D. Hoffman, Farhad Hormozdiari, Neil Houlsby, Shaobo Hou, Ghassen Jerfel, Alan Karthikesalingam, Mario Lucic, Yian Ma, Cory McLean, Diana Mincu, Akinori Mitani, Andrea Montanari, Zachary Nado, Vivek Natarajan, Christopher Nielson, Thomas F. Osborne, Rajiv Raman, Kim Ramasamy, Rory Sayres, Jessica Schrouff, Martin Seneviratne, Shannon Sequeira, Harini Suresh, Victor Veitch, Max Vladymyrov, Xuezhi Wang, Kellie Webster, Steve Yadlowsky, Taedong Yun, Xiaohua Zhai, D. Sculley
Links:
YouTube: https://www.youtube.com/c/yannickilcher
Twitter: https://twitter.com/ykilcher
Discord: https://discord.gg/4H8xxDF
BitChute: https://www.bitchute.com/channel/yannic-kilcher
Minds: https://www.minds.com/ykilcher
Parler: https://parler.com/profile/YannicKilcher
LinkedIn: https://www.linkedin.com/in/yannic-kilcher-488534136/
If you want to support me, the best thing to do is to share out the content :)
If you want to support me financially (completely optional and voluntary, but a lot of people have asked for this):
SubscribeStar: https://www.subscribestar.com/yannickilcher
Patreon: https://www.patreon.com/yannickilcher
Bitcoin (BTC): bc1q49lsw3q325tr58ygf8sudx2dqfguclvngvy2cq
Ethereum (ETH): 0x7ad3513E3B8f66799f507Aa7874b1B0eBC7F85e2
Litecoin (LTC): LQW2TRyKYetVC8WjFkhpPhtpbDM4Vw7r9m
Monero (XMR): 4ACL8AGrEo5hAir8A9CeVrW8pEauWvnp1WnSDZxW7tziCDLhZAGsgzhRQABDnFy8yuM9fWJDviJPHKRjV4FWt19CJZN9D4n
Видео Underspecification Presents Challenges for Credibility in Modern Machine Learning (Paper Explained) канала Yannic Kilcher
Deep Learning models are often overparameterized and have many degrees of freedom, which leads to many local minima that all perform equally well on the test set. But it turns out that even though they all generalize in-distribution, the performance of these models can be drastically different when tested out-of-distribution. Notably, in many cases, a good model can actually be found among all these candidates, but it seems impossible to select it. This paper describes this problem, which it calls underspecification, and gives several theoretical and practical examples.
OUTLINE:
0:00 - Into & Overview
2:00 - Underspecification of ML Pipelines
11:15 - Stress Tests
12:40 - Epidemiological Example
20:45 - Theoretical Model
26:55 - Example from Medical Genomics
34:00 - ImageNet-C Example
36:50 - BERT Models
56:55 - Conclusion & Comments
Paper: https://arxiv.org/abs/2011.03395
Abstract:
ML models often exhibit unexpectedly poor behavior when they are deployed in real-world domains. We identify underspecification as a key reason for these failures. An ML pipeline is underspecified when it can return many predictors with equivalently strong held-out performance in the training domain. Underspecification is common in modern ML pipelines, such as those based on deep learning. Predictors returned by underspecified pipelines are often treated as equivalent based on their training domain performance, but we show here that such predictors can behave very differently in deployment domains. This ambiguity can lead to instability and poor model behavior in practice, and is a distinct failure mode from previously identified issues arising from structural mismatch between training and deployment domains. We show that this problem appears in a wide variety of practical ML pipelines, using examples from computer vision, medical imaging, natural language processing, clinical risk prediction based on electronic health records, and medical genomics. Our results show the need to explicitly account for underspecification in modeling pipelines that are intended for real-world deployment in any domain.
Authors: Alexander D'Amour, Katherine Heller, Dan Moldovan, Ben Adlam, Babak Alipanahi, Alex Beutel, Christina Chen, Jonathan Deaton, Jacob Eisenstein, Matthew D. Hoffman, Farhad Hormozdiari, Neil Houlsby, Shaobo Hou, Ghassen Jerfel, Alan Karthikesalingam, Mario Lucic, Yian Ma, Cory McLean, Diana Mincu, Akinori Mitani, Andrea Montanari, Zachary Nado, Vivek Natarajan, Christopher Nielson, Thomas F. Osborne, Rajiv Raman, Kim Ramasamy, Rory Sayres, Jessica Schrouff, Martin Seneviratne, Shannon Sequeira, Harini Suresh, Victor Veitch, Max Vladymyrov, Xuezhi Wang, Kellie Webster, Steve Yadlowsky, Taedong Yun, Xiaohua Zhai, D. Sculley
Links:
YouTube: https://www.youtube.com/c/yannickilcher
Twitter: https://twitter.com/ykilcher
Discord: https://discord.gg/4H8xxDF
BitChute: https://www.bitchute.com/channel/yannic-kilcher
Minds: https://www.minds.com/ykilcher
Parler: https://parler.com/profile/YannicKilcher
LinkedIn: https://www.linkedin.com/in/yannic-kilcher-488534136/
If you want to support me, the best thing to do is to share out the content :)
If you want to support me financially (completely optional and voluntary, but a lot of people have asked for this):
SubscribeStar: https://www.subscribestar.com/yannickilcher
Patreon: https://www.patreon.com/yannickilcher
Bitcoin (BTC): bc1q49lsw3q325tr58ygf8sudx2dqfguclvngvy2cq
Ethereum (ETH): 0x7ad3513E3B8f66799f507Aa7874b1B0eBC7F85e2
Litecoin (LTC): LQW2TRyKYetVC8WjFkhpPhtpbDM4Vw7r9m
Monero (XMR): 4ACL8AGrEo5hAir8A9CeVrW8pEauWvnp1WnSDZxW7tziCDLhZAGsgzhRQABDnFy8yuM9fWJDviJPHKRjV4FWt19CJZN9D4n
Видео Underspecification Presents Challenges for Credibility in Modern Machine Learning (Paper Explained) канала Yannic Kilcher
Показать
Комментарии отсутствуют
Информация о видео
Другие видео канала
LambdaNetworks: Modeling long-range Interactions without Attention (Paper Explained)Rethinking Attention with Performers (Paper Explained)ReBeL - Combining Deep Reinforcement Learning and Search for Imperfect-Information Games (Explained)AI/ML: Challenges and Roadblocks of Machine LearningTraining more effective learned optimizers, and using them to train themselves (Paper Explained)AI VS ML VS DL VS Data ScienceHopfield Networks is All You Need (Paper Explained)Fourier Neural Operator for Parametric Partial Differential Equations (Paper Explained)Language Models are Open Knowledge Graphs (Paper Explained)Class 2: Artificial Intelligence, Machine Learning, and Deep LearningArtificial intelligence and algorithms: pros and cons | DW Documentary (AI documentary)OpenAI CLIP: ConnectingText and Images (Paper Explained)Learning To Classify Images Without Labels (Paper Explained)Nyströmformer: A Nyström-Based Algorithm for Approximating Self-Attention (AI Paper Explained)[Classic] Deep Residual Learning for Image Recognition (Paper Explained)DeepMind's AlphaFold 2 Explained! AI Breakthrough in Protein Folding! What we know (& what we don't)Feedback Transformers: Addressing Some Limitations of Transformers with Feedback Memory (Explained)The AI Economist: Improving Equality and Productivity with AI-Driven Tax Policies (Paper Explained)Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity