===Altruism, Selfishness, and Spite in Traffic Routing=== Po-An Chen\\ Department of Computer Science\\ Viterbi School of Engineering\\ University of Southern California\\ Wednesday, March 31st, 2010\\ 2:00pm-3:00pm\\ EEB 248\\ **Abstract:** We study the price of anarchy of traffic routing, under the assumption that users are partially altruistic or spiteful. We model such behavior by positing that the "cost" perceived by a user is a linear combination of the actual latency of the route chosen (selfish component), and the increase in latency the user causes for others (altruistic component). We show that if all users have a coefficient of at least \beta > 0 for the altruistic component,then the price of anarchy is bounded by 1/\beta, for all network topologies, arbitrary commodities, and arbitrary semi-convex latency functions. We extend this result to give more precise bounds on the price of anarchy for specific classes of latency functions, even for \beta < 0 modeling spiteful behavior. In particular, we show that if all latency functions are linear, the price of anarchy is bounded by 4/(3 + 2\beta - \beta^2). We next study non-uniform altruism distributions, where different users may have different coefficients \beta. We prove that all such games, even with infinitely many types of players, have a Nash Equilibrium. We show that if the average of the coefficients for the altruistic components of all users is \bar{\beta}, then the price of anarchy is bounded by 1/\bar{\beta}, for single commodity parallel link networks, and arbitrary convex latency functions. In particular, this result generalizes, albeit non-constructively, the Stackelberg routing results of Roughgarden and of Swamy. More generally, we bound the price of anarchy based on the class of allowable latency functions, and as a corollary obtain tighter bounds for Stackelberg routing than a recent result of Swamy. **Biography:** Po-An Chen received his B.B.A. degree and M.B.A degree both in Information Management from National Taiwan University in 2001 and 2003, respectively. He is a Ph.D. candidate in Computer Science at USC, advised by Prof. David Kempe. He also received M.S. degree in Computer Science from USC in 2007. His research interests are generally in theoretical computer science and algorithms, specifiically applications to algorithmic game theory, auctions and mechanisms design, distributed algorithms and multiagent systems. His current focus is the effect of altruism and spite in game-theoretic settings such as in routing, auctions, and network inoculation, as well as learning in repeated games. Advisor: Prof. David Kempe, dkempe [at] usc.edu Back to [[start | CommNetS Seminar Page ]]
