Applying Genetic Programming to Improve Interpretability in Machine Learning Models

TL;DR

This paper introduces GPX, a genetic programming-based method that generates interpretable local explanations for complex AI models, enhancing understanding of their decisions across various datasets.

Contribution

The paper presents a novel GP-based approach, GPX, for local interpretability of black-box models, outperforming existing methods in accuracy and clarity.

Findings

GPX produces more accurate explanations than current state-of-the-art methods.

The approach is effective across different complex models like RF, DNN, and SVM.

GPX generates comprehensible symbolic expressions reflecting local model behavior.

Abstract

Explainable Artificial Intelligence (or xAI) has become an important research topic in the fields of Machine Learning and Deep Learning. In this paper, we propose a Genetic Programming (GP) based approach, named Genetic Programming Explainer (GPX), to the problem of explaining decisions computed by AI systems. The method generates a noise set located in the neighborhood of the point of interest, whose prediction should be explained, and fits a local explanation model for the analyzed sample. The tree structure generated by GPX provides a comprehensible analytical, possibly non-linear, symbolic expression which reflects the local behavior of the complex model. We considered three machine learning techniques that can be recognized as complex black-box models: Random Forest, Deep Neural Network and Support Vector Machine in twenty data sets for regression and classifications problems. Our results indicate that the GPX is able to produce more accurate understanding of complex models than the state of the art. The results validate the proposed approach as a novel way to deploy GP to improve interpretability.