The operator approach to entropy games

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Entropy games and matrix multiplication games have been recently introduced by Asarin et al. They model the situation in which one player (Despot) wishes to minimize the growth rate of a matrix product, whereas the other player (Tribune) wishes to maximize it. We develop an operator approach to entropy games. This allows us to show that entropy games can be cast as stochastic mean payoff games in which some action spaces are simplices and payments are given by a relative entropy (Kullback-Leibler divergence). In this way, we show that entropy games with a fixed number of states belonging to Despot can be solved in polynomial time. This approach also allows us to solve these games by a policy iteration algorithm, which we compare with the spectral simplex algorithm developed by Protasov.

Original languageEnglish
Title of host publication34th Symposium on Theoretical Aspects of Computer Science, STACS 2017
EditorsBrigitte Vallee, Heribert Vollmer
PublisherSchloss Dagstuhl- Leibniz-Zentrum fur Informatik GmbH, Dagstuhl Publishing
ISBN (Electronic)9783959770286
DOIs
Publication statusPublished - 1 Mar 2017
Event34th Symposium on Theoretical Aspects of Computer Science, STACS 2017 - Hannover, Germany
Duration: 8 Mar 201711 Mar 2017

Publication series

NameLeibniz International Proceedings in Informatics, LIPIcs
Volume66
ISSN (Print)1868-8969

Conference

Conference34th Symposium on Theoretical Aspects of Computer Science, STACS 2017
Country/TerritoryGermany
CityHannover
Period8/03/1711/03/17

Keywords

  • Perron eigenvalues
  • Policy iteration
  • Risk sensitive control
  • Shapley operators
  • Stochastic games

Fingerprint

Dive into the research topics of 'The operator approach to entropy games'. Together they form a unique fingerprint.

Cite this