Partially observable stochastic game-based multi-agent prediction markets

Janyl Jumadinova, Prithviraj Dasgupta

Research output: Contribution to conferencePaper

2 Citations (Scopus)

Abstract

We present a novel representation of the prediction market using a partially observable stochastic game with information (POSGI), that can be used by each trading agent to precisely calculate the state of the market. We then propose that a correlated equilibrium (CE) strategy can be used by the agents to dynamically calculate the prices at which they should trade securities in the prediction market. Simulation results comparing the CE strategy within our POSGI model with five other strategies commonly used in similar markets show that the CE strategy results in improved price predictions and higher utilities to the agents as compared to other strategies.

Original languageEnglish (US)
Pages1149-1150
Number of pages2
StatePublished - Jan 1 2011
Event10th International Conference on Autonomous Agents and Multiagent Systems 2011, AAMAS 2011 - Taipei, Taiwan, Province of China
Duration: May 2 2011May 6 2011

Conference

Conference10th International Conference on Autonomous Agents and Multiagent Systems 2011, AAMAS 2011
CountryTaiwan, Province of China
CityTaipei
Period5/2/115/6/11

Keywords

  • Correlated equilibrium
  • Prediction market
  • Stochastic game

ASJC Scopus subject areas

  • Artificial Intelligence

Cite this

Jumadinova, J., & Dasgupta, P. (2011). Partially observable stochastic game-based multi-agent prediction markets. 1149-1150. Paper presented at 10th International Conference on Autonomous Agents and Multiagent Systems 2011, AAMAS 2011, Taipei, Taiwan, Province of China.

Partially observable stochastic game-based multi-agent prediction markets. / Jumadinova, Janyl; Dasgupta, Prithviraj.

2011. 1149-1150 Paper presented at 10th International Conference on Autonomous Agents and Multiagent Systems 2011, AAMAS 2011, Taipei, Taiwan, Province of China.

Research output: Contribution to conferencePaper

Jumadinova, J & Dasgupta, P 2011, 'Partially observable stochastic game-based multi-agent prediction markets', Paper presented at 10th International Conference on Autonomous Agents and Multiagent Systems 2011, AAMAS 2011, Taipei, Taiwan, Province of China, 5/2/11 - 5/6/11 pp. 1149-1150.
Jumadinova J, Dasgupta P. Partially observable stochastic game-based multi-agent prediction markets. 2011. Paper presented at 10th International Conference on Autonomous Agents and Multiagent Systems 2011, AAMAS 2011, Taipei, Taiwan, Province of China.
Jumadinova, Janyl ; Dasgupta, Prithviraj. / Partially observable stochastic game-based multi-agent prediction markets. Paper presented at 10th International Conference on Autonomous Agents and Multiagent Systems 2011, AAMAS 2011, Taipei, Taiwan, Province of China.2 p.
@conference{d082dc8d83ab444cb5cf60fc6e9756cb,
title = "Partially observable stochastic game-based multi-agent prediction markets",
abstract = "We present a novel representation of the prediction market using a partially observable stochastic game with information (POSGI), that can be used by each trading agent to precisely calculate the state of the market. We then propose that a correlated equilibrium (CE) strategy can be used by the agents to dynamically calculate the prices at which they should trade securities in the prediction market. Simulation results comparing the CE strategy within our POSGI model with five other strategies commonly used in similar markets show that the CE strategy results in improved price predictions and higher utilities to the agents as compared to other strategies.",
keywords = "Correlated equilibrium, Prediction market, Stochastic game",
author = "Janyl Jumadinova and Prithviraj Dasgupta",
year = "2011",
month = "1",
day = "1",
language = "English (US)",
pages = "1149--1150",
note = "10th International Conference on Autonomous Agents and Multiagent Systems 2011, AAMAS 2011 ; Conference date: 02-05-2011 Through 06-05-2011",

}

TY - CONF

T1 - Partially observable stochastic game-based multi-agent prediction markets

AU - Jumadinova, Janyl

AU - Dasgupta, Prithviraj

PY - 2011/1/1

Y1 - 2011/1/1

N2 - We present a novel representation of the prediction market using a partially observable stochastic game with information (POSGI), that can be used by each trading agent to precisely calculate the state of the market. We then propose that a correlated equilibrium (CE) strategy can be used by the agents to dynamically calculate the prices at which they should trade securities in the prediction market. Simulation results comparing the CE strategy within our POSGI model with five other strategies commonly used in similar markets show that the CE strategy results in improved price predictions and higher utilities to the agents as compared to other strategies.

AB - We present a novel representation of the prediction market using a partially observable stochastic game with information (POSGI), that can be used by each trading agent to precisely calculate the state of the market. We then propose that a correlated equilibrium (CE) strategy can be used by the agents to dynamically calculate the prices at which they should trade securities in the prediction market. Simulation results comparing the CE strategy within our POSGI model with five other strategies commonly used in similar markets show that the CE strategy results in improved price predictions and higher utilities to the agents as compared to other strategies.

KW - Correlated equilibrium

KW - Prediction market

KW - Stochastic game

UR - http://www.scopus.com/inward/record.url?scp=84899430717&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84899430717&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:84899430717

SP - 1149

EP - 1150

ER -