Fair and Interpretable Algorithmic Hiring using Evolutionary Many Objective Optimization

Hiring is a high-stakes decision-making process that balances the joint objectives of being fair and accurately selecting top candidates. The industry standard method employs subject-matter experts to manually generate hiring algorithms; however, this resource intensive finds sub-optimal solutions. Despite recognized need for algorithmic solutions address these limitations, no reported currently supports optimizing predictive while complying legal fairness standards. We present novel application Evolutionary Many-Objective Optimization (EMOO) methods create first fair, interpretable, legally compliant approach. Using proposed Dirichlet-based genetic operators improved search, we compare state-of-the-art EMOO models (NSGA-III, SPEA2-SDE, bi-goal evolution) expert solutions, verifying our results across three real world datasets diverse organizational positions. Experimental demonstrate outperform human experts, consistently fairer algorithms, can provide additional lift when removing constraints required analysis.