Reinforcement Learning Explained via Reinforcement Learning: Towards Explainable Policies through Predictive Explanation

Léo Saulières, Martin Cooper, Florence Bannay

2023

Abstract

In the context of reinforcement learning (RL), in order to increase trust in or understand the failings of an agent’s policy, we propose predictive explanations in the form of three scenarios: best-case, worst-case and most-probable. After showing W[1]-hardness of finding such scenarios, we propose linear-time approximations. In particular, to find an approximate worst/best-case scenario, we use RL to obtain policies of the environment viewed as a hostile/favorable agent. Experiments validate the accuracy of this approach.

Paper Citation

in Harvard Style

Saulières L., Cooper M. and Bannay F. (2023). Reinforcement Learning Explained via Reinforcement Learning: Towards Explainable Policies through Predictive Explanation. In Proceedings of the 15th International Conference on Agents and Artificial Intelligence - Volume 2: ICAART, ISBN 978-989-758-623-1, pages 35-44. DOI: 10.5220/0011619600003393

in Bibtex Style

@conference{icaart23,
author={Léo Saulières and Martin Cooper and Florence Bannay},
title={Reinforcement Learning Explained via Reinforcement Learning: Towards Explainable Policies through Predictive Explanation},
booktitle={Proceedings of the 15th International Conference on Agents and Artificial Intelligence - Volume 2: ICAART,},
year={2023},
pages={35-44},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0011619600003393},
isbn={978-989-758-623-1},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 15th International Conference on Agents and Artificial Intelligence - Volume 2: ICAART,
TI - Reinforcement Learning Explained via Reinforcement Learning: Towards Explainable Policies through Predictive Explanation
SN - 978-989-758-623-1
AU - Saulières L.
AU - Cooper M.
AU - Bannay F.
PY - 2023
SP - 35
EP - 44
DO - 10.5220/0011619600003393