Distributionally Robust Fair Principal Components via Geodesic Descents

Hieu Vu · Toan Tran · Man-Chung Yue · Viet Anh Nguyen

Keywords: [ distributionally robust optimization ]

[ Abstract ]
[ Visit Poster at Spot A0 in Virtual World ] [ OpenReview
Wed 27 Apr 2:30 a.m. PDT — 4:30 a.m. PDT


Principal component analysis is a simple yet useful dimensionality reduction technique in modern machine learning pipelines. In consequential domains such as college admission, healthcare and credit approval, it is imperative to take into account emerging criteria such as the fairness and the robustness of the learned projection. In this paper, we propose a distributionally robust optimization problem for principal component analysis which internalizes a fairness criterion in the objective function. The learned projection thus balances the trade-off between the total reconstruction error and the reconstruction error gap between subgroups, taken in the min-max sense over all distributions in a moment-based ambiguity set. The resulting optimization problem over the Stiefel manifold can be efficiently solved by a Riemannian subgradient descent algorithm with a sub-linear convergence rate. Our experimental results on real-world datasets show the merits of our proposed method over state-of-the-art baselines.

Chat is not available.