INFORMS
New Orleans, Nov. 13-16, 2005
AI Cluster 

Track: Multi-Agent System Applications
Chair: Riyaz Sikora, University of Texas at Arlington

Paper1: Foraging for Trust: Exploring Rationality and the Stag Hunt Game
Author: Steve O. Kimbrough, Wharton School of Business, Univ. of Pennsylvania,
Abstract: Trust presents a number of problems and paradoxes, because existing theory is not fully adequate for understanding why there is so much of it, why it occurs, and so forth. This paper explores the generation of trust with two simple, but very different models, focusing on repeated play of the Stag Hunt game. A gridscape model examines creation of trust among cognitively basic simple agents. A Markov model examines play between two somewhat more sophisticated agents.

Paper2: An Application for Discrimination with Strategic Behavior
Authors: Fidan Boylu, Haldun Aytug, and Gary Koehler, University of Florida
Abstract: Rational agents subject to classification by a principal might attempt to influence the classification as in loan decision situations where an agent might try to increase his chances of being accepted by altering his true attribute values. Exploring this strategic gaming, we apply our results to a credit-risk evaluation dataset.

Paper3: Learning Optimal Seller Policy: Application of Reinforcement and Evolutionary Learning
Authors: Riyaz Sikora and Vishal Sachdev, University of Texas at Arlington
Abstract: We consider the problem, recently reported in the literature, of homogeneous sellers of a single raw material or component vying for business from a single large buyer. Standard game-theoretic analysis of the problem assumes completely rational and omniscient agents to derive equilibrium seller policy. We relax those assumptions and present simple reinforcement and evolutionary learning agents that learn near-optimal seller policies.

Steve Kimbrough

Steve Kimbrough

Fidan Boylu

Riyaz Sikora