A multi-agent prediction market based on partially observable stochastic game

Janyl Jumadinova, Prithviraj Dasgupta

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

We present a novel, game theoretic representation called POSGI (partially observable stochastic game with information) for distributed information aggregation using a multiagent based prediction market model. We then describe a correlated equilibrium (CE)-based solution strategy for this game which enables each agent to dynamically calculate the prices at which it should trade a security in the prediction market. We have extended our results to risk averse traders and shown that a Pareto optimal correlated equilibrium strategy can be used to incentively truthful revelations from risk averse agents. Simulation results comparing our CE strategy with five other strategies commonly used in similar markets, with both risk neutral and risk averse agents, show that the CE strategy improves price predictions and provides higher utilities to the agents as compared to other existing strategies.

Original languageEnglish (US)
Title of host publicationProceedings of the 13th International Conference on Electronic Commerce, ICEC'11
DOIs
StatePublished - Dec 1 2011
Event13th International Conference on Electronic Commerce, ICEC'11 - Liverpool, United Kingdom
Duration: Aug 3 2011Aug 5 2011

Publication series

NameACM International Conference Proceeding Series

Conference

Conference13th International Conference on Electronic Commerce, ICEC'11
CountryUnited Kingdom
CityLiverpool
Period8/3/118/5/11

Fingerprint

Agglomeration

Keywords

  • Correlated equilibrium
  • Distributed information aggregation
  • Prediction market
  • Stochastic game

ASJC Scopus subject areas

  • Software
  • Human-Computer Interaction
  • Computer Vision and Pattern Recognition
  • Computer Networks and Communications

Cite this

Jumadinova, J., & Dasgupta, P. (2011). A multi-agent prediction market based on partially observable stochastic game. In Proceedings of the 13th International Conference on Electronic Commerce, ICEC'11 [2378125] (ACM International Conference Proceeding Series). https://doi.org/10.1145/2378104.2378125

A multi-agent prediction market based on partially observable stochastic game. / Jumadinova, Janyl; Dasgupta, Prithviraj.

Proceedings of the 13th International Conference on Electronic Commerce, ICEC'11. 2011. 2378125 (ACM International Conference Proceeding Series).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Jumadinova, J & Dasgupta, P 2011, A multi-agent prediction market based on partially observable stochastic game. in Proceedings of the 13th International Conference on Electronic Commerce, ICEC'11., 2378125, ACM International Conference Proceeding Series, 13th International Conference on Electronic Commerce, ICEC'11, Liverpool, United Kingdom, 8/3/11. https://doi.org/10.1145/2378104.2378125
Jumadinova J, Dasgupta P. A multi-agent prediction market based on partially observable stochastic game. In Proceedings of the 13th International Conference on Electronic Commerce, ICEC'11. 2011. 2378125. (ACM International Conference Proceeding Series). https://doi.org/10.1145/2378104.2378125
Jumadinova, Janyl ; Dasgupta, Prithviraj. / A multi-agent prediction market based on partially observable stochastic game. Proceedings of the 13th International Conference on Electronic Commerce, ICEC'11. 2011. (ACM International Conference Proceeding Series).
@inproceedings{3d23e205b7df4f079d0829dadb3973ab,
title = "A multi-agent prediction market based on partially observable stochastic game",
abstract = "We present a novel, game theoretic representation called POSGI (partially observable stochastic game with information) for distributed information aggregation using a multiagent based prediction market model. We then describe a correlated equilibrium (CE)-based solution strategy for this game which enables each agent to dynamically calculate the prices at which it should trade a security in the prediction market. We have extended our results to risk averse traders and shown that a Pareto optimal correlated equilibrium strategy can be used to incentively truthful revelations from risk averse agents. Simulation results comparing our CE strategy with five other strategies commonly used in similar markets, with both risk neutral and risk averse agents, show that the CE strategy improves price predictions and provides higher utilities to the agents as compared to other existing strategies.",
keywords = "Correlated equilibrium, Distributed information aggregation, Prediction market, Stochastic game",
author = "Janyl Jumadinova and Prithviraj Dasgupta",
year = "2011",
month = "12",
day = "1",
doi = "10.1145/2378104.2378125",
language = "English (US)",
isbn = "9781450314282",
series = "ACM International Conference Proceeding Series",
booktitle = "Proceedings of the 13th International Conference on Electronic Commerce, ICEC'11",

}

TY - GEN

T1 - A multi-agent prediction market based on partially observable stochastic game

AU - Jumadinova, Janyl

AU - Dasgupta, Prithviraj

PY - 2011/12/1

Y1 - 2011/12/1

N2 - We present a novel, game theoretic representation called POSGI (partially observable stochastic game with information) for distributed information aggregation using a multiagent based prediction market model. We then describe a correlated equilibrium (CE)-based solution strategy for this game which enables each agent to dynamically calculate the prices at which it should trade a security in the prediction market. We have extended our results to risk averse traders and shown that a Pareto optimal correlated equilibrium strategy can be used to incentively truthful revelations from risk averse agents. Simulation results comparing our CE strategy with five other strategies commonly used in similar markets, with both risk neutral and risk averse agents, show that the CE strategy improves price predictions and provides higher utilities to the agents as compared to other existing strategies.

AB - We present a novel, game theoretic representation called POSGI (partially observable stochastic game with information) for distributed information aggregation using a multiagent based prediction market model. We then describe a correlated equilibrium (CE)-based solution strategy for this game which enables each agent to dynamically calculate the prices at which it should trade a security in the prediction market. We have extended our results to risk averse traders and shown that a Pareto optimal correlated equilibrium strategy can be used to incentively truthful revelations from risk averse agents. Simulation results comparing our CE strategy with five other strategies commonly used in similar markets, with both risk neutral and risk averse agents, show that the CE strategy improves price predictions and provides higher utilities to the agents as compared to other existing strategies.

KW - Correlated equilibrium

KW - Distributed information aggregation

KW - Prediction market

KW - Stochastic game

UR - http://www.scopus.com/inward/record.url?scp=84867708459&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84867708459&partnerID=8YFLogxK

U2 - 10.1145/2378104.2378125

DO - 10.1145/2378104.2378125

M3 - Conference contribution

AN - SCOPUS:84867708459

SN - 9781450314282

T3 - ACM International Conference Proceeding Series

BT - Proceedings of the 13th International Conference on Electronic Commerce, ICEC'11

ER -