Actor-Critic Algorithms for Risk-Sensitive MDPs

Prashanth L.A. 1 Mohammad Ghavamzadeh 1
1 SEQUEL - Sequential Learning
LIFL - Laboratoire d'Informatique Fondamentale de Lille, LAGIS - Laboratoire d'Automatique, Génie Informatique et Signal, Inria Lille - Nord Europe
Abstract : In many sequential decision-making problems we may want to manage risk by minimizing some measure of variability in rewards in addition to maximizing a standard criterion. Variance-related risk measures are among the most common risk-sensitive criteria in finance and operations research. However, optimizing many such criteria is known to be a hard problem. In this paper, we consider both discounted and average reward Markov decision processes. For each formulation, we first define a measure of variability for a policy, which in turn gives us a set of risk-sensitive criteria to optimize. For each of these criteria, we derive a formula for computing its gradient. We then devise actor-critic algorithms for estimating the gradient and updating the policy parameters in the ascent direction. We establish the convergence of our algorithms to locally risk-sensitive optimal policies. Finally, we demonstrate the usefulness of our algorithms in a traffic signal control application.
Document type :
Reports
Complete list of metadatas

Cited literature [38 references]  Display  Hide  Download

https://hal.inria.fr/hal-00794721
Contributor : Mohammad Ghavamzadeh <>
Submitted on : Wednesday, October 16, 2013 - 4:44:35 AM
Last modification on : Thursday, February 21, 2019 - 10:52:49 AM
Long-term archiving on : Friday, April 7, 2017 - 11:33:44 AM

File

rs-rl-techreport.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-00794721, version 2

Collections

Citation

Prashanth L.A., Mohammad Ghavamzadeh. Actor-Critic Algorithms for Risk-Sensitive MDPs. [Technical Report] 2013. ⟨hal-00794721v2⟩

Share

Metrics

Record views

419

Files downloads

563