Solving Factored MDPs with Hybrid State and Action Variables
Abstract
Efficient representations and solutions for large decision problems with continuous and discrete variables are among the most important challenges faced by the designers of automated decision support systems. In this paper, we describe a novel hybrid factored Markov decision process (MDP) model that allows for a compact representation of these problems, and a new hybrid approximate linear programming (HALP) framework that permits their efficient solutions. The central idea of HALP is to approximate the optimal value function by a linear combination of basis functions and optimize its weights by linear programming. We analyze both theoretical and computational aspects of this approach, and demonstrate its scale-up potential on several hybrid optimization problems.
Cite
Text
Kveton et al. "Solving Factored MDPs with Hybrid State and Action Variables." Journal of Artificial Intelligence Research, 2006. doi:10.1613/JAIR.2085Markdown
[Kveton et al. "Solving Factored MDPs with Hybrid State and Action Variables." Journal of Artificial Intelligence Research, 2006.](https://mlanthology.org/jair/2006/kveton2006jair-solving/) doi:10.1613/JAIR.2085BibTeX
@article{kveton2006jair-solving,
title = {{Solving Factored MDPs with Hybrid State and Action Variables}},
author = {Kveton, Branislav and Hauskrecht, Milos and Guestrin, Carlos},
journal = {Journal of Artificial Intelligence Research},
year = {2006},
pages = {153-201},
doi = {10.1613/JAIR.2085},
volume = {27},
url = {https://mlanthology.org/jair/2006/kveton2006jair-solving/}
}