On Markov Games Played by Bayesian and Boundedly-Rational Players

Muthukumaran Chandrasekaran; Yingke Chen; Prashant Doshi

On Markov Games Played by Bayesian and Boundedly-Rational Players

Muthukumaran Chandrasekaran
, Yingke Chen
, Prashant Doshi

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

101 Downloads (Pure)

Abstract

We present a new game-theoretic framework in which
Bayesian players with bounded rationality engage in a
Markov game and each has private but incomplete information
regarding other players’ types. Instead of utilizing
Harsanyi’s abstract types and a common prior, we construct
intentional player types whose structure is explicit and induces
a finite-level belief hierarchy. We characterize an equilibrium
in this game and establish the conditions for existence
of the equilibrium. The computation of finding such equilibria
is formalized as a constraint satisfaction problem and its
effectiveness is demonstrated on two cooperative domains.

Original language	English
Title of host publication	Proceedings of 31st AAAI Conference on Artificial Intelligence
Publisher	AAAI
Number of pages	7
Publication status	Published - 13 Feb 2017
Event	31st AAAI Conference on Artificial Intelligence - San Francisco, United States Duration: 4 Feb 2017 → 9 Feb 2017

Publication series

Name	Proceedings of the AAAI conference on artificial intelligence
ISSN (Electronic)	2374-3468

Conference

Conference	31st AAAI Conference on Artificial Intelligence
Abbreviated title	AAAI-17
Country/Territory	United States
City	San Francisco
Period	4/02/17 → 9/02/17

Bibliographical note

We present a new game-theoretic framework in which Bayesian players with bounded rationality engage in a Markov game and each has private but incomplete information regarding other players’ types. Instead of utilizing Harsanyi’s abstract types and a common prior, we construct intentional player types whose structure is explicit and induces a finite-level belief hierarchy. We characterize an equilibrium in this game and establish the conditions for existence of the equilibrium. The computation of finding such equilibria is formalized as a constraint satisfaction problem and its effectiveness is demonstrated on two cooperative domains.

Access to Document

Full textAccepted author manuscript, 3.77 MBLicence: CC BY-NC-ND

https://aaai.org/ocs/index.php/AAAI/AAAI17/paper/view/14334/13795Licence: CC BY-NC-ND

Cite this

@inproceedings{839f304787524f08872f9912c5233866,

title = "On Markov Games Played by Bayesian and Boundedly-Rational Players",

abstract = "We present a new game-theoretic framework in which Bayesian players with bounded rationality engage in a Markov game and each has private but incomplete information regarding other players{\textquoteright} types. Instead of utilizing Harsanyi{\textquoteright}s abstract types and a common prior, we construct intentional player types whose structure is explicit and induces a finite-level belief hierarchy. We characterize an equilibrium in this game and establish the conditions for existence of the equilibrium. The computation of finding such equilibria is formalized as a constraint satisfaction problem and its effectiveness is demonstrated on two cooperative domains.",

author = "Muthukumaran Chandrasekaran and Yingke Chen and Prashant Doshi",

note = "We present a new game-theoretic framework in which Bayesian players with bounded rationality engage in a Markov game and each has private but incomplete information regarding other players{\textquoteright} types. Instead of utilizing Harsanyi{\textquoteright}s abstract types and a common prior, we construct intentional player types whose structure is explicit and induces a finite-level belief hierarchy. We characterize an equilibrium in this game and establish the conditions for existence of the equilibrium. The computation of finding such equilibria is formalized as a constraint satisfaction problem and its effectiveness is demonstrated on two cooperative domains.; 31st AAAI Conference on Artificial Intelligence, AAAI-17 ; Conference date: 04-02-2017 Through 09-02-2017",

year = "2017",

month = feb,

day = "13",

language = "English",

series = " Proceedings of the AAAI conference on artificial intelligence",

publisher = "AAAI",

booktitle = "Proceedings of 31st AAAI Conference on Artificial Intelligence",

}

Chandrasekaran, M, Chen, Y & Doshi, P 2017, On Markov Games Played by Bayesian and Boundedly-Rational Players. in Proceedings of 31st AAAI Conference on Artificial Intelligence. Proceedings of the AAAI conference on artificial intelligence, AAAI, 31st AAAI Conference on Artificial Intelligence, San Francisco, United States, 4/02/17. <https://aaai.org/ocs/index.php/AAAI/AAAI17/paper/view/14334/13795>

TY - GEN

T1 - On Markov Games Played by Bayesian and Boundedly-Rational Players

AU - Chandrasekaran, Muthukumaran

AU - Chen, Yingke

AU - Doshi, Prashant

N1 - We present a new game-theoretic framework in which Bayesian players with bounded rationality engage in a Markov game and each has private but incomplete information regarding other players’ types. Instead of utilizing Harsanyi’s abstract types and a common prior, we construct intentional player types whose structure is explicit and induces a finite-level belief hierarchy. We characterize an equilibrium in this game and establish the conditions for existence of the equilibrium. The computation of finding such equilibria is formalized as a constraint satisfaction problem and its effectiveness is demonstrated on two cooperative domains.

PY - 2017/2/13

Y1 - 2017/2/13

N2 - We present a new game-theoretic framework in which Bayesian players with bounded rationality engage in a Markov game and each has private but incomplete information regarding other players’ types. Instead of utilizing Harsanyi’s abstract types and a common prior, we construct intentional player types whose structure is explicit and induces a finite-level belief hierarchy. We characterize an equilibrium in this game and establish the conditions for existence of the equilibrium. The computation of finding such equilibria is formalized as a constraint satisfaction problem and its effectiveness is demonstrated on two cooperative domains.

AB - We present a new game-theoretic framework in which Bayesian players with bounded rationality engage in a Markov game and each has private but incomplete information regarding other players’ types. Instead of utilizing Harsanyi’s abstract types and a common prior, we construct intentional player types whose structure is explicit and induces a finite-level belief hierarchy. We characterize an equilibrium in this game and establish the conditions for existence of the equilibrium. The computation of finding such equilibria is formalized as a constraint satisfaction problem and its effectiveness is demonstrated on two cooperative domains.

M3 - Conference contribution

T3 - Proceedings of the AAAI conference on artificial intelligence

BT - Proceedings of 31st AAAI Conference on Artificial Intelligence

PB - AAAI

T2 - 31st AAAI Conference on Artificial Intelligence

Y2 - 4 February 2017 through 9 February 2017

ER -

On Markov Games Played by Bayesian and Boundedly-Rational Players

Abstract

Publication series

Conference

Bibliographical note

Access to Document

Fingerprint

Cite this