Timezone: »

Calibrated Ensembles: A Simple Way to Mitigate ID-OOD Accuracy Tradeoffs
Ananya Kumar · Aditi Raghunathan · Tengyu Ma · Percy Liang
Event URL: https://openreview.net/forum?id=dmDE-9e9F_x »

We often see undesirable tradeoffs in robust machine learning where out-of-distribution (OOD) accuracy is at odds with in-distribution (ID) accuracy. A ‘robust’ classifier obtained via specialized techniques like removing spurious features has better OOD but worse ID accuracy compared to a ‘standard’ classifier trained via vanilla ERM. On six distribution shift datasets, we find that simply ensembling the standard and robust models is a strong baseline---we match the ID accuracy of a standard model with only a small drop in OOD accuracy compared to the robust model. However, calibrating these models in-domain surprisingly improves the OOD accuracy of the ensemble and completely eliminates the tradeoff and we achieve the best of both ID and OOD accuracy over the original models.

Author Information

Ananya Kumar (Stanford University)
Aditi Raghunathan (Stanford University)
Tengyu Ma (Stanford University)
Percy Liang (Stanford University)
Percy Liang

Percy Liang is an Assistant Professor of Computer Science at Stanford University (B.S. from MIT, 2004; Ph.D. from UC Berkeley, 2011). His research spans machine learning and natural language processing, with the goal of developing trustworthy agents that can communicate effectively with people and improve over time through interaction. Specific topics include question answering, dialogue, program induction, interactive learning, and reliable machine learning. His awards include the IJCAI Computers and Thought Award (2016), an NSF CAREER Award (2016), a Sloan Research Fellowship (2015), and a Microsoft Research Faculty Fellowship (2014).

More from the Same Authors