English

A proof of convergence of inverse reinforcement learning for multi-objective optimization

Machine Learning 2023-05-19 v3 Artificial Intelligence Machine Learning

Abstract

We show the convergence of Wasserstein inverse reinforcement learning for multi-objective optimizations with the projective subgradient method by formulating an inverse problem of the multi-objective optimization problem. In addition, we prove convergence of inverse reinforcement learning (maximum entropy inverse reinforcement learning, guided cost learning) with gradient descent and the projective subgradient method.

Keywords

Cite

@article{arxiv.2305.06137,
  title  = {A proof of convergence of inverse reinforcement learning for multi-objective optimization},
  author = {Akira Kitaoka and Riki Eto},
  journal= {arXiv preprint arXiv:2305.06137},
  year   = {2023}
}

Comments

10 pages