Deriving equivalent symbol-based decision models from feedforward neural networks

TL;DR

This paper introduces a method to convert opaque neural networks into interpretable decision trees, making AI systems more transparent and trustworthy.

Contribution

A systematic approach to derive symbolic decision models from feedforward neural networks, enhancing interpretability and trust in AI.

Findings

Symbolic models like decision trees can be derived from FNNs while preserving functionality.

A prototype using Keras and TensorFlow demonstrates the feasibility of the proposed method.

The approach enables scalability to deeper networks through iterative refinement of subpaths.

Abstract

Artificial intelligence (AI) has emerged as a transformative force across industries, driven by advances in deep learning and natural language processing, and fueled by large-scale data and computing resources. Despite its rapid adoption, the opacity of AI systems poses significant challenges to trust and acceptance. This work explores the intersection of connectionist and symbolic approaches to artificial intelligence, focusing on the derivation of interpretable symbolic models, such as decision trees, from feedforward neural networks (FNNs). Decision trees provide a transparent framework for elucidating the operations of neural networks while preserving their functionality. The derivation is presented in a step-by-step approach and illustrated with several examples. A systematic methodology is proposed to bridge neural and symbolic paradigms by exploiting distributed representations in FNNs to identify symbolic components, including fillers, roles, and their interrelationships. The process traces neuron activation values and input configurations across network layers, mapping activations and their underlying inputs to decision tree edges. The resulting symbolic structures effectively capture FNN decision processes and enable scalability to deeper networks through iterative refinement of subpaths for each hidden layer. To validate the theoretical framework, a prototype was developed using Keras .h5-data and emulating TensorFlow within the Java JDK/JavaFX environment. This prototype demonstrates the feasibility of extracting symbolic representations from neural networks, enhancing trust in AI systems, and promoting accountability.

Figures