Agent Interactions and Implicit Trust in IPD Environments
- 978 Downloads
The goal of multi-agent systems is to build robust intelligent systems capable of existing in complex environments. Agents must decide with whom to interact. In this paper we investigate how agents may bias their interactions in environments where alternative game payoffs are available. We present a number of game theoretic simulations involving a range of agent interaction models. Through a series of experiments we show the effects of modelling agent interactions when games representing alternative levels of benefit and risk are offered. Individual agents may have a preference for games of a certain risk. We also present analysis of population dynamics, examining how agents bias their peer interactions throughout each generation. We also address the topic of implicit trust, where agents reflect levels of trust through the payoffs presented in a game offer. In this interaction model agents may use levels of trust to choose opponents and to determine levels of risk associated with a game.
KeywordsInteraction Model Cooperative Strategy Agent Interaction Repeated Interaction Game Interaction
Unable to display preview. Download preview PDF.
- 1.Axelrod, R.: The Evolution of Cooperation. Basic Books, New York (1984)Google Scholar
- 5.Holland, J.: The effects of labels (tags) on social interactions. Working Paper, Santa Fe Institute 93-10-064 (1993)Google Scholar
- 6.Howley, E., O’Riordan, C.: Agent cooperation using simple fixed bias tags and multiple tags. In: AICS 2005. The 16th Irish Conference On Artificial Intelligence and Cognative Science (2005)Google Scholar
- 7.Howley, E., O’Riordan, C.: The emergence of cooperation among agents using simple fixed bias tagging. In: IEEE CEC 2005. Proceedings of the 2005 Congress on Evolutionary Computation, vol. 2, pp. 1011–1016. IEEE Press, Los Alamitos (2005)Google Scholar
- 8.Howley, E., O’Riordan, C.: The effects and evolution of implicit trust in populations playing the iterated prisoner’s dilemma. In: Proceedings of the 2006 Congress on Evolutionary Computation IEEE CEC 2006. Held as part of the IEEE World Congress On Computational Intelligence IEEE WCCI 2006. IEEE Press, Los Alamitos (2006)Google Scholar
- 9.Howley, E., O’Riordan, C.: The effects of viscosity in choice and refusal ipd environments. In: Bell, D.A., Milligan, P., Sage, P.P. (eds.) Procs. of the Seventeenth Irish Conference on Artificial Intelligence and Cognitive Science, Queen’s University Belfast, pp. 213–222 (2006)Google Scholar
- 10.Marsh, S.: Trust in distributed artificial intelligence. In: Castelfranchi, C., Werner, E. (eds.) MAAMAW 1992. LNCS, vol. 830, pp. 94–112. Springer, Heidelberg (1994)Google Scholar
- 11.Marsh, S.: Formalising trust as a computational concept. Ph.D. Thesis, University of Stirling (1994), http://citeseer.ist.psu.edu/marsh94formalising.html
- 13.Riolo, R.: The effects and evolution of tag-mediated selection of partners in populations playing the iterated prisoner’s dilemma. In: ICGA, pp. 378–385 (1997)Google Scholar
- 14.Stanley, E.A., Ashlock, D., Smucker, M.D.: Iterated prisoner’s dilemma with choice and refusal of partners: Evolutionary results. In: Moran, F., Merelo, J.J., Moreno, A., Chacon, P. (eds.) Advances in Artificial Life. LNCS, vol. 929, pp. 490–502. Springer, Heidelberg (1995)Google Scholar
- 16.Turner, H., Kazakov, D.: Stochastic simulation of inherited kinship-driven altruism. Journal of Artificial Intelligence and Simulation of Behaviour 1(2) (2002)Google Scholar