Group-realizable multi-group learning by minimizing empirical risk

Navid Ardeshir, Samuel Deng, Daniel Hsu, Jingwen Liu

Keywords: Multi-group learning, group-realizability, empirical risk minimization, sample complexity, computational complexity

TL;DR: Sample complexity of multi-group learning is shown to improve in the group-realizable setting over the agnostic setting via ERM

Abstract: The sample complexity of multi-group learning is shown to improve in the group-realizable setting over the agnostic setting, even when the family of groups is infinite so long as it has finite VC dimension. The improved sample complexity is obtained by empirical risk minimization over the class of group-realizable concepts, which itself could have infinite VC dimension. Implementing this approach is also shown to be computationally intractable, and an alternative approach is suggested based on improper learning.

Submission Number: 91