Distribution-Free Statistical Dispersion Control for Societal Applications
Abstract
Explicit finite-sample statistical guarantees on model performance are an important ingredient in responsible machine learning. Previous work has focused mainly on bounding either the expected loss of a predictor or the probability that an individual prediction will incur a loss value in a specified range. However, for many high-stakes applications, it is crucial to understand and control the dispersion of a loss distribution, or the extent to which different members of a population experience unequal effects of algorithmic decisions. We initiate the study of distribution-free control of statistical dispersion measures with societal implications and propose a simple yet flexible framework that allows us to handle a much richer class of statistical functionals beyond previous work. Our methods are verified through experiments in toxic comment detection, medical imaging, and film recommendation.
Cite
@article{arxiv.2309.13786,
title = {Distribution-Free Statistical Dispersion Control for Societal Applications},
author = {Zhun Deng and Thomas P. Zollo and Jake C. Snell and Toniann Pitassi and Richard Zemel},
journal= {arXiv preprint arXiv:2309.13786},
year = {2024}
}
Comments
Accepted by NeurIPS as spotlight (top 3% among submissions)