Reinforcement Learning with Parameterized Actions
Artificial Intelligence
2015-11-30 v4 Machine Learning
Abstract
We introduce a model-free algorithm for learning in Markov decision processes with parameterized actions-discrete actions with continuous parameters. At each step the agent must select both which action to use and which parameters to use with that action. We introduce the Q-PAMDP algorithm for learning in these domains, show that it converges to a local optimum, and compare it to direct policy search in the goal-scoring and Platform domains.
Cite
@article{arxiv.1509.01644,
title = {Reinforcement Learning with Parameterized Actions},
author = {Warwick Masson and Pravesh Ranchod and George Konidaris},
journal= {arXiv preprint arXiv:1509.01644},
year = {2015}
}
Comments
Accepted for AAAI 2016