FA(IR)2MA-GLVQ – A hidden-feature-bias mitigation approach for fairness in classification learning based on generalized matrix learning vector quantization

2026/03/02 by Marika Kaden, Ronny Schubert, Julius Voigt +6 · 1 voice
Social Sciences · Computer Science · #Ethics and Social Impacts of AI #Adversarial Robustness in Machine Learning #Domain Adaptation and Few-Shot Learning

paper · doi:10.1016/j.neucom.2026.133200

openalex publication_date 2026/03/02 · openalex created_date 2026/03/03 · openalex updated_date 2026/07/02

Abstract

Developing fair classification models is a crucial aspect of machine learning research. However, unintended distortion in training data - biased data - can lead to discriminatory decisions. In this paper, we developed a workflow for detecting and mitigating bias in data using a shallow, interpretable machine learning models: the Generalized Matrix Learning Vector Quantization. We extent the approach by a relevance-based analysis to identify and reduce bias in the data. Combining similarity metric adaptation and relevance-based analysis, we can develop fair classification models that minimize the influence of bias in the data. Our results demonstrate that this method is effective in reducing bias in classification models and therefore supports fair decision-making.

Citations

• Interpretable Classification Models for Recidivism Prediction
• On a Test of Whether one of Two Random Variables is Stochastically Larger than the Other
• Approximate Statistical Tests for Comparing Supervised Classification Learning Algorithms
• Improving imbalanced learning through a heuristic oversampling method based on k-means and SMOTE
• Amazing Things Come From Having Many Good Models
• Iterated Relevance Matrix Analysis (IRMA) for the identification of class-discriminative subspaces
• The Rashomon Importance Distribution: Getting RID of Unstable, Single Model-based Variable Importance
• An Empirical Evaluation of the Rashomon Effect in Explainable Machine Learning
• Do highly over-parameterized neural networks generalize since bad solutions are rare?
• Stability and Generalization Analysis of Gradient Methods for Shallow Neural Networks
• Bias Mitigation for Machine Learning Classifiers: A Comprehensive Survey
• Provably Adversarially Robust Nearest Prototype Classifiers
• Interpretable Machine Learning: Fundamental Principles and 10 Grand Challenges
• Machine learning for precision medicine
• Null It Out: Guarding Protected Attributes by Iterative Nullspace\n Projection
• Explainable Artificial Intelligence (XAI): Concepts, Taxonomies, Opportunities and Challenges toward Responsible AI
• A Survey on Bias and Fairness in Machine Learning
• A Unified Approach to Interpreting Model Predictions
• "Why Should I Trust You?": Explaining the Predictions of Any Classifier

Discussions

• The importance of lunch breaks cannot be overrated! This work emerged from a discussion at ESANN2024, was presented at ESANN2025 and extended for Neurocomputing. With @hsmw.bsky.social colleagues, @ [bsky, 1 points, 0 comments]

Related

• Bias Mitigation for Machine Learning Classifiers: A Comprehensive Survey
• Bias-Tolerant Fair Classification
• Two-stage Algorithm for Fairness-aware Machine Learning
• Identifying and Correcting Label Bias in Machine Learning
• Metrics and methods for a systematic comparison of fairness-aware machine learning algorithms