Using Fuzzy Logic for Improving Model Interpretability in Machine Learning Classifiers

Abstract

This research explores the integration of fuzzy logic into traditional machine learning classifiers to enhance model interpretability without significantly sacrificing predictive accuracy. As machine learning models increasingly influence critical decisions in fields like healthcare, finance, and autonomous systems, the need for transparent and understandable decision-making processes has become paramount. This study compares a traditional machine learning model with a fuzzy-enhanced version, evaluating their performance based on accuracy, fidelity, simplicity, and stability. While the fuzzy-enhanced model shows a slight reduction in accuracy (84.9% compared to 85.7% for the traditional model), it offers substantial improvements in interpretability and consistency. The fuzzy model achieves high fidelity (92.5%), uses a simplified decision-making process with fewer rules and a shallower tree depth, and demonstrates greater stability in its explanations. These findings suggest that incorporating fuzzy logic into machine learning classifiers can create models that are not only effective but also more transparent and trustworthy, making them better suited for applications where interpretability is critical.