English

KL Divergence Estimation with Multi-group Attribution

Machine Learning 2022-03-01 v1 Information Theory math.IT Machine Learning

Abstract

Estimating the Kullback-Leibler (KL) divergence between two distributions given samples from them is well-studied in machine learning and information theory. Motivated by considerations of multi-group fairness, we seek KL divergence estimates that accurately reflect the contributions of sub-populations to the overall divergence. We model the sub-populations coming from a rich (possibly infinite) family C\mathcal{C} of overlapping subsets of the domain. We propose the notion of multi-group attribution for C\mathcal{C}, which requires that the estimated divergence conditioned on every sub-population in C\mathcal{C} satisfies some natural accuracy and fairness desiderata, such as ensuring that sub-populations where the model predicts significant divergence do diverge significantly in the two distributions. Our main technical contribution is to show that multi-group attribution can be derived from the recently introduced notion of multi-calibration for importance weights [HKRR18, GRSW21]. We provide experimental evidence to support our theoretical results, and show that multi-group attribution provides better KL divergence estimates when conditioned on sub-populations than other popular algorithms.

Keywords

Cite

@article{arxiv.2202.13576,
  title  = {KL Divergence Estimation with Multi-group Attribution},
  author = {Parikshit Gopalan and Nina Narodytska and Omer Reingold and Vatsal Sharan and Udi Wieder},
  journal= {arXiv preprint arXiv:2202.13576},
  year   = {2022}
}

Comments

20 pages, 4 figures

R2 v1 2026-06-24T09:55:49.718Z