🤖 AI Summary
Existing fair classification methods often rely on strong distributional assumptions and require large sample sizes, rendering them unable to rigorously guarantee group fairness—especially in few-shot settings (e.g., <500 samples). To address this, we propose FaiREE, the first algorithm that provides **rigorous theoretical guarantees** for mainstream fairness constraints—including Equality of Opportunity—under **no distributional assumptions** and for **arbitrary finite sample sizes**. At its core, FaiREE integrates empirical risk minimization with a conservative, confidence-based fairness boundary, forming a risk-bounded fair optimization framework. Theoretical analysis establishes that FaiREE achieves optimal classification accuracy while provably satisfying fairness constraints. Extensive experiments on synthetic and real-world datasets demonstrate that FaiREE significantly outperforms state-of-the-art methods, maintaining stable fairness compliance even in low-sample regimes; these empirical results robustly validate its theoretical guarantees.
📝 Abstract
Algorithmic fairness plays an increasingly critical role in machine learning research. Several group fairness notions and algorithms have been proposed. However, the fairness guarantee of existing fair classification methods mainly depends on specific data distributional assumptions, often requiring large sample sizes, and fairness could be violated when there is a modest number of samples, which is often the case in practice. In this paper, we propose FaiREE, a fair classification algorithm that can satisfy group fairness constraints with finite-sample and distribution-free theoretical guarantees. FaiREE can be adapted to satisfy various group fairness notions (e.g., Equality of Opportunity, Equalized Odds, Demographic Parity, etc.) and achieve the optimal accuracy. These theoretical guarantees are further supported by experiments on both synthetic and real data. FaiREE is shown to have favorable performance over state-of-the-art algorithms.
