Distilling Deep RL Models Into Interpretable Neuro-Fuzzy Systems

TL;DR

This paper introduces a method to convert deep reinforcement learning policies into compact, interpretable neuro-fuzzy controllers, maintaining high performance with significantly fewer rules.

Contribution

The authors develop a distillation algorithm that transforms deep Q-network policies into small neuro-fuzzy controllers, enhancing interpretability without sacrificing performance.

Findings

Nearly match DQN performance with only 2-6 fuzzy rules

Effective in three OpenAI Gym environments

Enables interpretable RL solutions

Abstract

Deep Reinforcement Learning uses a deep neural network to encode a policy, which achieves very good performance in a wide range of applications but is widely regarded as a black box model. A more interpretable alternative to deep networks is given by neuro-fuzzy controllers. Unfortunately, neuro-fuzzy controllers often need a large number of rules to solve relatively simple tasks, making them difficult to interpret. In this work, we present an algorithm to distill the policy from a deep Q-network into a compact neuro-fuzzy controller. This allows us to train compact neuro-fuzzy controllers through distillation to solve tasks that they are unable to solve directly, combining the flexibility of deep reinforcement learning and the interpretability of compact rule bases. We demonstrate the algorithm on three well-known environments from OpenAI Gym, where we nearly match the performance of a DQN agent using only 2 to 6 fuzzy rules.