Timezone: »

 
Domain-Adjusted Regression or: ERM May Already Learn Features Sufficient for Out-of-Distribution Generalization
Elan Rosenfeld · Pradeep Ravikumar · Andrej Risteski
Event URL: https://openreview.net/forum?id=Ypo0AckYW8 »

A common explanation for the failure of deep networks to generalize out-of-distribution is that they fail to recover the "correct" features. We challenge this notion with a simple experiment which suggests that ERM already learns sufficient features and that the current bottleneck is not feature learning, but robust regression. We therefore argue that devising simpler methods for learning predictors on existing features is a promising direction for future research. Towards this end, we introduce Domain-Adjusted Regression (DARE), a convex objective for learning a linear predictor that is provably robust under a new model of distribution shift. Rather than learning one function, DARE performs a domain-specific adjustment to unify the domains in a canonical latent space and learns to predict in this space. Under a natural model, we prove that the DARE solution is the minimax-optimal predictor for a constrained set of test distributions. Further, we provide the first finite-environment convergence guarantee to the minimax risk, improving over existing analyses which only yield minimax predictors after an environment threshold. Evaluated on finetuned features, we find that DARE compares favorably to prior methods, consistently achieving equal or better performance.

Author Information

Elan Rosenfeld (Carnegie Mellon University)
Pradeep Ravikumar (Carnegie Mellon University)
Andrej Risteski (CMU)

Assistant Professor in the ML department at CMU. Prior to that I was a Wiener Fellow at MIT, and prior to that finished my PhD at Princeton University.

Related Events (a corresponding poster, oral, or spotlight)

More from the Same Authors