Optimizing Quantiles in Preference-Based Markov Decision Processes

Authors

  • Hugo Gilbert Pierre and Marie Curie University
  • Paul Weng Sun Yat-sen University
  • Yan Xu Carnegie Mellon University

DOI:

https://doi.org/10.1609/aaai.v31i1.11026

Keywords:

Markov decision process, Quantile

Abstract

In the Markov decision process model, policies are usually evaluated by expected cumulative rewards. As this decision criterion is not always suitable, we propose in this paper an algorithm for computing a policy optimal for the quantile criterion. Both finite and infinite horizons are considered. Finally we experimentally evaluate our approach on random MDPs and on a data center control problem.

AAAI-2017 Proceedings Cover

Downloads

Published

2017-02-12

How to Cite

Gilbert, H., Weng, P., & Xu, Y. (2017). Optimizing Quantiles in Preference-Based Markov Decision Processes. Proceedings of the AAAI Conference on Artificial Intelligence, 31(1). https://doi.org/10.1609/aaai.v31i1.11026

Issue

Vol. 31 No. 1 (2017): Thirty-First AAAI Conference on Artificial Intelligence

Section

Main Track: Planning and Scheduling