Unleashing Linear Optimizers for Group-Fair Learning and Optimization

Abstract

Most systems and learning algorithms optimize average performance or average loss -- one reason being computational complexity. However, many objectives of practical interest are more complex than simply average loss. This arises, for example, when balancing performance or loss with fairness across people. We prove that, from a computational perspective, optimizing arbitrary objectives that take into account performance over a small number of groups is not significantly harder to optimize than average performance. Our main result is a polynomial-time reduction that uses a linear optimizer to optimize an arbitrary (Lipschitz continuous) function of performance over a (constant) number of possibly-overlapping groups. This includes fairness objectives over small numbers of groups, and we further point out that other existing notions of fairness such as individual fairness can be cast as convex optimization and hence more standard convex techniques can be used. Beyond learning, our approach applies to multi-objective optimization, more generally.

arXiv commentAccepted for presentation at the Conference on Learning Theory (COLT) 2018

Cited by in corpus (10)

• Fairness in Machine Learning: A Survey
• Fairness in Machine Learning
• Fair Regression: Quantitative Definitions and Reduction-based Algorithms
• Fairness risk measures
• Ensuring Fairness Beyond the Training Data
• Rényi Fair Inference
• Optimizing Black-box Metrics with Adaptive Surrogates
• Doubly-stochastic mining for heterogeneous retrieval
• Optimizing Generalized Rate Metrics through Game Equilibrium
• The Cost of a Reductions Approach to Private Fair Optimization