Learning Fair Representations

TLDR

The study proposes a learning algorithm that simultaneously achieves group and individual fairness in classification. The algorithm learns a data representation by optimizing an objective that balances accurate encoding with obfuscation of protected group membership. Experiments on three datasets show the method outperforms existing techniques and enables transfer learning and metric learning for classification.

Abstract

We propose a learning algorithm for fair classification that achieves both group fairness (the proportion of members in a protected group receiving positive classification is identical to the proportion in the population as a whole), and individual fairness (similar individuals should be treated similarly). We formulate fairness as an optimization problem of finding a good representation of the data with two competing goals: to encode the data as well as possible, while simultaneously obfuscating any information about membership in the protected group. We show positive results of our algorithm relative to other known techniques, on three datasets. Moreover, we demonstrate several advantages to our approach. First, our intermediate representation can be used for other classification tasks (i.e., transfer learning is possible); secondly, we take a step toward learning a distance metric which can find important dimensions of the data for classification.