[ICML 2020] Consistent Estimators for Learning to Defer to an Expert

ICML 2020

[ICML 2020] Consistent Estimators for Learning to Defer to an Expert

Jul 16, 2020
|
42 views
|
|
Code
Details
ICML 2020 video presentation on our paper "Consistent Estimators for Learning to Defer to an Expert" by Hussein Mozannar and David Sontag. Abstract: Learning algorithms are often used in conjunction with expert decision makers in practical scenarios, however this fact is largely ignored when designing these algorithms. In this paper we explore how to learn predictors that can either predict or choose to defer the decision to a downstream expert. Given only samples of the expert's decisions, we give a procedure based on learning a classifier and a rejector and analyze it theoretically. Our approach is based on a novel reduction to cost sensitive learning where we give a consistent surrogate loss for cost sensitive learning that generalizes the cross entropy loss. We show the effectiveness of our approach on a variety of experimental tasks. paper link: https://arxiv.org/abs/2006.01862 code: https://github.com/clinicalml/learn-to-defer

Comments
loading...