Symbolic Explanation of Affinity-Based Reinforcement Learning Agents with Markov Models

TL;DR

This paper introduces a method to interpret reinforcement learning policies by using symbolic Markov models that capture the global affinities of learned strategies, demonstrated in personalized financial decision-making.

Contribution

It presents a novel policy regularization technique that makes reinforcement learning policies inherently interpretable through symbolic Markov models.

Findings

Successfully explained policies in personalized prosperity management

Demonstrated the interpretability of strategies via discretized Markov models

Enhanced understanding of reinforcement learning behaviors

Abstract

The proliferation of artificial intelligence is increasingly dependent on model understanding. Understanding demands both an interpretation - a human reasoning about a model's behavior - and an explanation - a symbolic representation of the functioning of the model. Notwithstanding the imperative of transparency for safety, trust, and acceptance, the opacity of state-of-the-art reinforcement learning algorithms conceals the rudiments of their learned strategies. We have developed a policy regularization method that asserts the global intrinsic affinities of learned strategies. These affinities provide a means of reasoning about a policy's behavior, thus making it inherently interpretable. We have demonstrated our method in personalized prosperity management where individuals' spending behavior in time dictate their investment strategies, i.e. distinct spending personalities may have dissimilar associations with different investment classes. We now explain our model by reproducing the underlying prototypical policies with discretized Markov models. These global surrogates are symbolic representations of the prototypical policies.