Toward Computationally Efficient Inverse Reinforcement Learning via Reward Shaping
Machine Learning
2023-12-19 v2 Artificial Intelligence
Machine Learning
Abstract
Inverse reinforcement learning (IRL) is computationally challenging, with common approaches requiring the solution of multiple reinforcement learning (RL) sub-problems. This work motivates the use of potential-based reward shaping to reduce the computational burden of each RL sub-problem. This work serves as a proof-of-concept and we hope will inspire future developments towards computationally efficient IRL.
Cite
@article{arxiv.2312.09983,
title = {Toward Computationally Efficient Inverse Reinforcement Learning via Reward Shaping},
author = {Lauren H. Cooke and Harvey Klyne and Edwin Zhang and Cassidy Laidlaw and Milind Tambe and Finale Doshi-Velez},
journal= {arXiv preprint arXiv:2312.09983},
year = {2023}
}