Change is Hard: A Closer Look at Subpopulation Shift
Jul 24, 2023
Speakers
Yuzhe Yang
Speaker · 0 followers
Haoran Zhang
Speaker · 0 followers
Dina Katabi
Speaker · 1 follower
Marzyeh Ghassemi
Speaker · 0 followers
About
Machine learning models often perform poorly on subgroups that are underrepresented in the training data. Yet, little is understood on the variation in mechanisms that cause subpopulation shifts, and how algorithms generalize across such diverse shifts at scale. In this work, we provide a fine-grained analysis of subpopulation shift. We first propose a unified framework that dissects and explains common shifts in subgroups. We then establish a comprehensive benchmark of 20 state-of-the-art algorithms evaluated on 12 real-world datasets in vision, language, and healthcare domains. With results obtained from training over 10,000 models, we reveal intriguing observations for future progress in this space. First, existing algorithms only improve subgroup robustness over certain types of shifts but not others. Moreover, while current algorithms rely on group-annotated validation data for model selection, we find that a simple selection criterion based on worst-class accuracy is surprisingly effective even without any group information. Finally, unlike existing works that solely aim to improve worst-group accuracy (WGA), we demonstrate the fundamental tradeoff between WGA and other important metrics, highlighting the need to carefully choose testing metrics. Code and data are available at: https://github.com/YyzHarry/SubpopBench.Machine learning models often perform poorly on subgroups that are underrepresented in the training data. Yet, little is understood on the variation in mechanisms that cause subpopulation shifts, and how algorithms generalize across such diverse shifts at scale. In this work, we provide a fine-grained analysis of subpopulation shift. We first propose a unified framework that dissects and explains common shifts in subgroups. We then establish a comprehensive benchmark of 20 state-of-the-art algo…
Like the format? Trust SlidesLive to capture your next event!
Professional recording and live streaming, delivered globally.
Sharing
Recommended Videos
Presentations on similar topic, category or speaker
Cluster-Specific Predictions with Multi-Task Gaussian Processes
Watch later
Arthur Leroy, …
A closer look at few-shot classification again
Watch later
Xu Luo, …
Pretraining Language Models with Human Preferences
Watch later
Tomek Korbak, …
Flipping Coins to Estimate Pseudocounts for Exploration in Reinforcement Learning
Watch later
Sam Lobel, …
