Optimising Equal Opportunity Fairness in Model Training

2022/05/05 by Aili Shen, Xudong Han, Shen, Aili +7 · 2 citations
Computer Science · Social Sciences · #Adversarial Robustness in Machine Learning #Computation and Language (cs.CL) #Ethics and Social Impacts of AI #Explainable Artificial Intelligence (XAI) #FOS: Computer and information sciences #Machine Learning (cs.LG)

paper · pdf · doi:10.48550/arxiv.2205.02393

openalex publication_date 2022/05/05 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Real-world datasets often encode stereotypes and societal biases. Such biases can be implicitly captured by trained models, leading to biased predictions and exacerbating existing societal preconceptions. Existing debiasing methods, such as adversarial training and removing protected information from representations, have been shown to reduce bias. However, a disconnect between fairness criteria and training objectives makes it difficult to reason theoretically about the effectiveness of different techniques. In this work, we propose two novel training objectives which directly optimise for the widely-used criterion of \it equal opportunity, and show that they are effective in reducing bias while maintaining high performance over two classification tasks.

Cited by

• Understanding Fairness and Prediction Error through Subspace Decomposition and Influence Analysis
• Benchmarking Gender and Political Bias in Large Language Models

Related

• Recovering from Biased Data: Can Fairness Constraints Improve Accuracy?
• Towards Equal Opportunity Fairness through Adversarial Learning
• Systematic Evaluation of Predictive Fairness
• Fair Infinitesimal Jackknife: Mitigating the Influence of Biased Training Data Points Without Refitting
• Coping with Mistreatment in Fair Algorithms