AI Training Series: Machine Learning Fairness

Anne Tumlin is a 2nd year PhD student in Computer Science at Vanderbilt University, where she works in Dr. Taylor T. Johnson’s Verification and Validation for Intelligent and Trustworthy Autonomy Laboratory (VeriVITAL) and the Network and Data Science (NDS) Lab with Dr. Tyler Derr. Her research focuses on AI and power systems, specifically analyzing the robustness of machine learning used in smart grids, while her passion project explores fairness in machine learning. Anne is a recipient of the DOE Computational Science Graduate Fellowship (CSGF) and the Provost Graduate Scholarship at Vanderbilt. She earned a BS in Computer Science from the University of South Carolina and will be interning next summer at Sandia National Labs in Albuquerque, NM.

Abstract: Ensuring fairness in machine learning (ML) is vital, especially as these models are increasingly used in socially critical financial decision-making processes such as credit scoring, loan approvals, and fraud detection. Fairness verification aims to provide formal guarantees of fairness in ML models. In this work, we introduce FairNNV, a tool that leverages the Neural Network Verification (NNV) framework to verify individual and counterfactual fairness using reachability analysis techniques. FairNNV introduces the Verified Fairness (VF) score to quantify fairness. Additionally, we compare the verification process of models before and after applying adversarial debiasing techniques to assess the impact of bias mitigation. We demonstrate FairNNV’s effectiveness on several fairness benchmark datasets, including Adult Census, German Credit, and Bank Marketing, with a focused analysis on the impact of adversarial debiasing on Adult Census classifiers. Experimental results show differences between empirical fairness improvements using adversarial debiasing and fairness verification scores with FairNNV, indicating a need for integrating formal verification into the evaluation process to guide model selections when assessing fairness.