Decision support using deterministic equivalents of probabilistic game trees

Michael L. Valenzuela, Liana Suantak, Jerzy W Rozenblit

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

3 Citations (Scopus)

Abstract

We have developed a game-theory driven decision-support tool that builds probabilistic game trees automatically from user-defined actions, rules, and states. The result of evaluating the paths in the game tree is a series of decisions which forms a decision-path representing an ε-Nash-Equilibrium. The algorithm uses certainty-equivalents to handle trade-offs between expected rewards and risks, effectively modeling the probabilistic game tree as deterministic. The resulting decision-paths correspond to player actions in the scenario. These sets of actions can be used as search patterns against a real-world database. A match to one of these patterns indicates an instance of novel behavior patterns generated by the game-theory driven decision support tool. This particular paradigm could be applied in any domain that requires anticipating and responding to adversarial agents with uncertainty, from mission planning to emergency responders to systems configuration.

Original languageEnglish (US)
Title of host publicationProceedings - 2012 IEEE 19th International Conference and Workshops on Engineering of Computer-Based Systems, ECBS 2012
Pages142-149
Number of pages8
DOIs
StatePublished - 2012
Event2012 IEEE 19th International Conference and Workshops on Engineering of Computer-Based Systems, ECBS 2012 - Novi Sad, Serbia
Duration: Apr 11 2012Apr 13 2012

Other

Other2012 IEEE 19th International Conference and Workshops on Engineering of Computer-Based Systems, ECBS 2012
CountrySerbia
CityNovi Sad
Period4/11/124/13/12

Fingerprint

Game theory
Planning
Uncertainty

Keywords

  • certainty equivalents
  • decision support
  • game simulation
  • repeated game theory
  • risk aversion

ASJC Scopus subject areas

  • Hardware and Architecture
  • Control and Systems Engineering

Cite this

Valenzuela, M. L., Suantak, L., & Rozenblit, J. W. (2012). Decision support using deterministic equivalents of probabilistic game trees. In Proceedings - 2012 IEEE 19th International Conference and Workshops on Engineering of Computer-Based Systems, ECBS 2012 (pp. 142-149). [6195180] https://doi.org/10.1109/ECBS.2012.22

Decision support using deterministic equivalents of probabilistic game trees. / Valenzuela, Michael L.; Suantak, Liana; Rozenblit, Jerzy W.

Proceedings - 2012 IEEE 19th International Conference and Workshops on Engineering of Computer-Based Systems, ECBS 2012. 2012. p. 142-149 6195180.

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

Valenzuela, ML, Suantak, L & Rozenblit, JW 2012, Decision support using deterministic equivalents of probabilistic game trees. in Proceedings - 2012 IEEE 19th International Conference and Workshops on Engineering of Computer-Based Systems, ECBS 2012., 6195180, pp. 142-149, 2012 IEEE 19th International Conference and Workshops on Engineering of Computer-Based Systems, ECBS 2012, Novi Sad, Serbia, 4/11/12. https://doi.org/10.1109/ECBS.2012.22
Valenzuela ML, Suantak L, Rozenblit JW. Decision support using deterministic equivalents of probabilistic game trees. In Proceedings - 2012 IEEE 19th International Conference and Workshops on Engineering of Computer-Based Systems, ECBS 2012. 2012. p. 142-149. 6195180 https://doi.org/10.1109/ECBS.2012.22
Valenzuela, Michael L. ; Suantak, Liana ; Rozenblit, Jerzy W. / Decision support using deterministic equivalents of probabilistic game trees. Proceedings - 2012 IEEE 19th International Conference and Workshops on Engineering of Computer-Based Systems, ECBS 2012. 2012. pp. 142-149
@inproceedings{d9d8d1cafb9f40f1904417455f02a950,
title = "Decision support using deterministic equivalents of probabilistic game trees",
abstract = "We have developed a game-theory driven decision-support tool that builds probabilistic game trees automatically from user-defined actions, rules, and states. The result of evaluating the paths in the game tree is a series of decisions which forms a decision-path representing an ε-Nash-Equilibrium. The algorithm uses certainty-equivalents to handle trade-offs between expected rewards and risks, effectively modeling the probabilistic game tree as deterministic. The resulting decision-paths correspond to player actions in the scenario. These sets of actions can be used as search patterns against a real-world database. A match to one of these patterns indicates an instance of novel behavior patterns generated by the game-theory driven decision support tool. This particular paradigm could be applied in any domain that requires anticipating and responding to adversarial agents with uncertainty, from mission planning to emergency responders to systems configuration.",
keywords = "certainty equivalents, decision support, game simulation, repeated game theory, risk aversion",
author = "Valenzuela, {Michael L.} and Liana Suantak and Rozenblit, {Jerzy W}",
year = "2012",
doi = "10.1109/ECBS.2012.22",
language = "English (US)",
isbn = "9780769546643",
pages = "142--149",
booktitle = "Proceedings - 2012 IEEE 19th International Conference and Workshops on Engineering of Computer-Based Systems, ECBS 2012",

}

TY - GEN

T1 - Decision support using deterministic equivalents of probabilistic game trees

AU - Valenzuela, Michael L.

AU - Suantak, Liana

AU - Rozenblit, Jerzy W

PY - 2012

Y1 - 2012

N2 - We have developed a game-theory driven decision-support tool that builds probabilistic game trees automatically from user-defined actions, rules, and states. The result of evaluating the paths in the game tree is a series of decisions which forms a decision-path representing an ε-Nash-Equilibrium. The algorithm uses certainty-equivalents to handle trade-offs between expected rewards and risks, effectively modeling the probabilistic game tree as deterministic. The resulting decision-paths correspond to player actions in the scenario. These sets of actions can be used as search patterns against a real-world database. A match to one of these patterns indicates an instance of novel behavior patterns generated by the game-theory driven decision support tool. This particular paradigm could be applied in any domain that requires anticipating and responding to adversarial agents with uncertainty, from mission planning to emergency responders to systems configuration.

AB - We have developed a game-theory driven decision-support tool that builds probabilistic game trees automatically from user-defined actions, rules, and states. The result of evaluating the paths in the game tree is a series of decisions which forms a decision-path representing an ε-Nash-Equilibrium. The algorithm uses certainty-equivalents to handle trade-offs between expected rewards and risks, effectively modeling the probabilistic game tree as deterministic. The resulting decision-paths correspond to player actions in the scenario. These sets of actions can be used as search patterns against a real-world database. A match to one of these patterns indicates an instance of novel behavior patterns generated by the game-theory driven decision support tool. This particular paradigm could be applied in any domain that requires anticipating and responding to adversarial agents with uncertainty, from mission planning to emergency responders to systems configuration.

KW - certainty equivalents

KW - decision support

KW - game simulation

KW - repeated game theory

KW - risk aversion

UR - http://www.scopus.com/inward/record.url?scp=84862084392&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84862084392&partnerID=8YFLogxK

U2 - 10.1109/ECBS.2012.22

DO - 10.1109/ECBS.2012.22

M3 - Conference contribution

AN - SCOPUS:84862084392

SN - 9780769546643

SP - 142

EP - 149

BT - Proceedings - 2012 IEEE 19th International Conference and Workshops on Engineering of Computer-Based Systems, ECBS 2012

ER -