Fairness with Exponential Weights

TL;DR

This paper introduces a meta-algorithm that transforms existing Hedge algorithms into fair contextual bandit algorithms, ensuring statistical parity and handling non-stationary environments with theoretical guarantees.

Contribution

It presents a novel meta-algorithm that guarantees exact statistical parity in contextual bandits by converting Hedge algorithms, accommodating non-stationarity and enabling batch classification with fairness.

Findings

Achieves statistical parity on every trial.

Maintains asymptotic regret bounds similar to Exp4.

Handles non-stationary distributions effectively.

Abstract

Motivated by the need to remove discrimination in certain applications, we develop a meta-algorithm that can convert any efficient implementation of an instance of Hedge (or equivalently, an algorithm for discrete bayesian inference) into an efficient algorithm for the equivalent contextual bandit problem which guarantees exact statistical parity on every trial. Relative to any comparator with statistical parity, the resulting algorithm has the same asymptotic regret bound as running the corresponding instance of Exp4 for each protected characteristic independently. Given that our Hedge instance admits non-stationarity we can handle a varying distribution with which to enforce statistical parity with respect to, which is useful when the true population is unknown and needs to be estimated from the data received so far. Via online-to-batch conversion we can handle the equivalent batch classification problem with exact statistical parity, giving us results that we believe are novel and important in their own right.