J. W. Weibull. Evolutionary Game Theory. MIT Press, 1997. A Missing Proofs Proof of Lemma 3.3: Let G = (S = f1; 2g; u 1 ; : : : ; u n ) be a symmetric game, p a uniform solution to G's basic linear system, and B = fS i (j)g, A = fS(j)g the basic and auxiliary sets, respectively. The vector p can be extended to a correlated equilibrium of G if and only if there is a solution q : S  Home/Search   Document Not in Database   Summary   Related Articles   Check

....rationally or to have complete knowledge of the details of the game. They act instead according to a pre programmed behavior pattern, or pure strategy, and it is supposed that some evolutionary selection process operates over time on the distribution of behaviors. We refer the reader to [8] [16] for excellent introductions to this rapidly expanding field. 9999 99 be the set of available pure strategies and, for all , let be the proportion of population members playing strategy , at time . The state of the population at a given instant is the vector ....

.... M1 A (3) where a dot signifies derivative with respect to time, and M 850 2 M is a function with open domain containing . Here, the function ( N ) specifies the rate at which pure strategy replicates. It is usually required that the growth function is regular [16], which means that it is Lipschitz continuous and that M A O . The former condition guarantees us that the system of differential equations (3) has a unique solution through any initial population state. The condition M A , instead, ensures that the simplex is invariant ....

[Article contains additional citation context not shown here]

J. W. Weibull, Evolutionary Game Theory. Cambridge, MA: MIT Press, 1995. 60

....of minority. 2.2 No information Nash equilibria If a minority game is played once, the Game Theory allows the study of the best choices to be taken for agents. When all agents behave in such a way that if they deviate from their current behavior they win less, a Nash equilibrium is attained [17]. For a minority game, the only rational behavior consists of taking random decisions. Of course, this is not very satisfactory and one has to give information to agents in order to obtain a non trivial behavior. 2.3 Information If the game is repeated, at least one agent may think that past ....

....total losses of agents. Therefore, Eq (4.10) implies that agents minimize their losses under this learning dynamics. In terms of Evolutionary Game theory, this means that they reach a Nash equilibrium. This is con rmed by the direct application of the multipopulation standard replicator dynamics [17] (RD) which is known to lead to a Nash equilibrium and reads = i;s a i;s ( j a j ) i a i ) j a j ) 4.11) Apart from the factor i , this coincides with the RD of Eq. 4.11) Again is minimized along the trajectories of Eq. 4.9) it is easy to check that the ....

J. W. Weibull, \Evolutionary Game Theory", MIT Press, Cambridge (1995)

....recent generalization of the Motzkin Straus theorem [19] to formulate the maximum clique problem as a quadratic programming problem. To (approximately) solve it we employ payoff monotonic dynamics, a class of simple dynamical systems recently developed and studied in evolutionary game theory [14] [23]. Such continuous solutions to discrete problems are interesting as they can motivate analog and biological implementations. It is worth remarking that traditional energyminimization graph matching algorithms such as [12] are not applicable to the tree matching problem (either rooted or unrooted) ....

....rationally or to have complete knowledge of the details of the game. They act instead according to a preprogrammed behavior pattern, or pure strategy, and it is supposed that some evolutionary selection process operates over time on the distribution of behaviors. We refer the reader to [14] [23] for excellent introductions to this rapidly expanding field. Let J f1; ###;ng be the set of available pure strategies and, for all i J , let x i be the proportion of population members playing strategy i, at time t. The state of the population at a given instant is the vector x ....

[Article contains additional citation context not shown here]

J.W. Weibull, Evolutionary Game Theory. Cambridge, Mass.: MIT Press, 1995.

....more natural interpretation as solving an optimization problem, the EGT analysis also provides some insight into the behavior of CCEAs as optimizers. We begin with some background discussion regarding evolutionary game theory, and introduce the concept of multi population symmetric games (MPS) [5], the subclass of EGT models that will be used to model and analyze these kinds of cooperative coevolutionary algorithms. We then describe the particular CCEAs under study and show how MPS can be used to model them. In section four we present our initial MPS analysis, introduce the idea of using ....

....terms of the specifics of what is learned, as well as the general message regarding the applicability of the MPS game framework. II. EVOLUTIONARY GAME THEORY EGT describes a set of dynamical systems models for which modern dynamical systems theory can be used to analyze evolutionary processes [5], 7] 8] and, as we will see, coevolutionary algorithms fit very nicely into this game theoretic framework. In evolutionary game theory, we are working with models of populations of individuals who are interacting with one another. Individuals repeatedly meet and receive some reward or ....

J. Weibull. Evolutionary game theory. MIT Press, Cambridge, MA, 1992.

....up to date bibliography of these applications. The model and its properties. In this section we discuss the basic intuition behind replicator equations and present a few theoretical properties that are instrumental for their application to optimization problems. For a more systematic treatment see [23, 55]. Consider a large, ideally infinite population of individuals belonging to the same species which compete for a particular limited resource, such as food, territory, etc. This kind of conflict is modeled as a game, the players being pairs of randomly selected population members. In contrast to ....

.... 0 we have tF(x(t) 0 for system (2) and F(x(t At) F(x(t) for system (3) unless x(t) is a stationary point. Furthermore, any such trajectory converges to a (unique) station ary point. The previous result is known in mathematical biology as the fundamental theorem of natural selection [17, 23, 55] and, in its original form, traces back to Fisher [18] As far as the discrete time model is concerned, it can be regarded as straightforward implication of the Baum Eagon theorem [2, 3] which is valid for general polyno mial functions over product of simplives. Waugh and Westervelt [54] also ....

[Article contains additional citation context not shown here]

WEIBULL, J. W.: Evolutionary Game Theory, MIT Press, Cambridge, MA, 1995.

....approach to economic simulation that will make up for the scarcity of data, while o#ering a scientific approach to data collection and a systematic tool for experimentation. The methodology we use is based on the evolutionary approach to game theoretical problems. Evolutionary game theory [29] studies equilibria of games played by populations of players, where players are myopically rational and have conflicting interests. In evolutionary systems there is no fitness function, instead there is a rule which governs survival of society members based on their success. So, the fitness of ....

J. W. Weibull. Evolutionary Game Theory. The MIT Press, 1995.

....a guide to behavior. An agent s optimal actions may be quite different depending upon whether it is playing against a perfectly rational agent or a boundedly rational person. This divergence led to the use of evolutionary methods for studying the bargaining behavior of boundedly rational agents [18, 9, 4, 17, 1, 3]. Although for certain games the game theoretic and evolutionary equilibria coincide [17, 16] in general, it has been shown that the game theoretic outcome may not always be valid when playing against boundedly rational agents [2] For instance, 18] and [4] show this in their evolutionary model ....

....against a perfectly rational agent or a boundedly rational person. This divergence led to the use of evolutionary methods for studying the bargaining behavior of boundedly rational agents [18, 9, 4, 17, 1, 3] Although for certain games the game theoretic and evolutionary equilibria coincide [17, 16], in general, it has been shown that the game theoretic outcome may not always be valid when playing against boundedly rational agents [2] For instance, 18] and [4] show this in their evolutionary model for the Nash demand game, as do Binmore et al. [1] in their evolutionary analysis of ....

[Article contains additional citation context not shown here]

J. W. Weibull. Evolutionary game theory. MIT Press, 1997.

....in a mixed strategy are cast as proportions of a large population of agents playing the corresponding pure strategies, then an agent population that has reached a fixed point with respect to the replicator dynamics will be a candidate symmetric mixed strategy Nash equilibrium. Weibull [29] shows that for 2 player games, fixed points of a broad class of replicator processes are Nash equilibria if neither strategy is extinct. For N player games, the set of fixed points that are locally asymptotically stable (all states su#ciently close converge to the same state) are a subset of the ....

J. W. Weibull. Evolutionary Game Theory. MIT Press, 1995.

....users diverting flow from routes with high delay to routes with low delay, though features in our algorithm such as time outs have no validity in the tra#c context. It is worth noting that this work has some kinship with the extensive literature on learning in evolutionary games [10] 23] [27], 25] and [28] These works, however, study repeated games, a much simpler situation than a full fledged stochastic dynamic game. Nevertheless, good convergence results are available only for some very simple cases. Our results provide a scheme for an apparently much more complex situation. It ....

....Nevertheless, good convergence results are available only for some very simple cases. Our results provide a scheme for an apparently much more complex situation. It may also be noted that the ordinary di#erential equation (ODE) limit (15) of our scheme is in fact the replicator dynamics of [27]. See also [21] 22] for related developments. For earlier routing algorithms which do not, however, have a learning component, see [3] 24] Our algorithm can also be regarded as a game theoretic learning problem with several interesting features. The most important is that each agent (node) ....

[Article contains additional citation context not shown here]

J. Weibull, Evolutionary Game Theory, MIT Press, Cambridge, MA 1995.

.... of possible states (and thus taken into consideration when choosing the next move) Recently, promising attempts have been made, especially in the field of evolutionary game theory, to use e.g. adaptive dynamics to describe how equilibria might be reached among simple strategies in iterated games [4, 8, 9]. While these attempts try to answer the question How are strategies behaving , we will here try to focus on the questions What is the result of their behavior and How can this result be used at the meta level . To help us answer these questions we will use Characteristic Distributions or ....

J. Weibull. Evolutionary Game Theory. MIT Press, 1996.

....or simply for other reasons have to interact. 1.2.7 Evolutionary Game Theory Evolutionary game theory deals with the problem of finding good and stable equilibria in evolutionary games. There have been several titles released during the last years that treats this matter, e.g. by Weibull [91], Samuelson [80] and Hofbauer and Sigmund [40] Through different mathematical tools such as replicator dynamics, analysis through Markov models and adaptive dynamics [68] the behavior of a population of rational agents may be traced, from its initial state, to an equilibrium (if it ....

.... of possible states (and thus taken in to consideration when choosing the next move) Recently, promising attempts have been made, especially in the field of evolutionary game theory, to use e.g. adaptive dynamics to describe how equilibria might be reached among simple strategies in iterated games [40, 80, 91]. While these attempts try to answer the question How are strategies behaving , we will here try to focus on the question What is the result of their behavior and How can this result be used . We have in previous chapters 3 4[22, 47] briefly discussed the notion of Characteristic ....

J. Weibull. Evolutionary Game Theory. MIT Press, 1996.

.... of mixed strategies and equilibria are discussed thoroughly for instance in the classic book by Maynard Smith [ 9 ] Rosenschein and Zlotkin formulated several agent scenarios in terms of game theory in their Rules of Encounter [ 11 ] and Weibull gives an rationalistic economics perspective [ 19 ] . Given that every system possible to exploit will be exploited, we must ask ourselves the question whether the behaviors described above (the law of Jante and the self con dent) are exploitable or not. 6.4.3.1 Exploiting the Janteists It would actually be enough not to underestimate the ....

J. Weibull. Evolutionary Game Theory. MIT Press, 1996.

....state dominated by those specied agents which in the course of the dynamiced evolution are able to establish a much larger number of links than the average agent. The paradigm to study the emergence of cooperation has been the Prisoner s Dilemma (PD) game. Using evolutionary game theory [18], it was shown [3,2] that cooperation may be sustained by a population of agents meeting repeatedly through globed random interactions. Two agents interact playing the game and, according to their outcome, their strategies are allowed to evolve. A second route to cooperative behavior, pioneered by ....

....result, which indicates that, in the long run, the searching capability of the D agents rewards them. This behavior is observed systematically in the parameter regime 1 b 2 for K = 8. The above result seems surprising from the point of view of the traditional repli cation dynamics [18] used in evolutionary game theory, because one could conclude from Fig. 2 that D agents should dominate the whole population. But our results indicate that the final highly cooperative state is not determined by average agents in the system, but rather by a small subset of those maximally ....

J. Weibull. Evolutionary Game Theory. MIT University Press, 1996.

....equation. The canonical replicator used in EGT is a difference (or differential) equation that selects agents to reproduce offspring in direct proportion to fitness. The dynamics of various replicator equations is a topic of intense study, particularly with respect to the gametheoretic equilibria [Weibull, 1995, Samuelson, 1997] Further, ties between EGT, quantitative genetics, and animal behavior have launched a host of more biologically inspired investigations into the dynamics of replicator systems [Hofbauer and Sigmund, 1998, Dugatkin and Reeve, 1998] Evolutionary game theory has also lead to a ....

Weibull, J. (1995). Evolutionary Game Theory. MIT Press.

No context found.

J. W. Weibull. Evolutionary Game Theory. MIT Press, 1997. A Missing Proofs Proof of Lemma 3.3: Let G = (S = f1; 2g; u 1 ; : : : ; u n ) be a symmetric game, p a uniform solution to G's basic linear system, and B = fS i (j)g, A = fS(j)g the basic and auxiliary sets, respectively. The vector p can be extended to a correlated equilibrium of G if and only if there is a solution q : S

No context found.

J.W.Weibull,Evolutionary Game Theory. Cambridge: MIT Press, 1995. 1011

No context found.

Jorgen W. Weibull. Evolutionary Game Theory. MIT press, 1995.

No context found.

J. W. Weibull. Evolutionary Game Theory. MIT press, 1995.

No context found.

J. W. Weibull. Evolutionary Game Theory. MIT Press 1997. 22

No context found.

WEIBULL, J. (1996): Evolutionary Game Theory. MIT University Press.

No context found.

Weibull, J.W., Evolutionary Game Theory, MIT Press, (1996).

No context found.

Jorgen W. Weibull. Evolutionary Game Theory. MIT press, 1995.

No context found.

Weibull, Jorgen W. Evolutionary Game Theory. The MIT Press, 1995.

No context found.

J. Weibull. Evolutionary Game Theory. The MIT Press, Cambridge, 1995.

No context found.

J. Weibull, Evolutionary Game Theory," MIT Press, Cambridge, MA, 1997.

No context found.

Weibull, J.W., Evolutionary Game Theory, MIT Press, (1996).

No context found.

Jorgen W. Weibull. Evolutionary Game Theory. MIT Press, Cambridge, Massachusetts, 1995.

No context found.

J orgen W. Weibull. Evolutionary Game Theory. MIT Press, 1995.

No context found.

J. W. Weibull. Evolutionary Game Theory. MIT Press, 1997. A Missing Proofs Proof of Lemma 3.3: Let G = (S = f1; 2g; u 1 ; : : : ; u n ) be a symmetric game, p a uniform solution to G's basic linear system, and B = fS i (j)g, A = fS(j)g the basic and auxiliary sets, respectively. The vector p can be extended to a correlated equilibrium of G if and only if there is a solution q : S

No context found.

Weibull, J. W. (1995), Evolutionary Game Theory, MIT Press, Cambridge.

No context found.

Weibull, J.W., Evolutionary Game Theory, MIT Press, (1996).

No context found.

Weibull, J. W., 1995. Evolutionary game theory, The MIT Press, Cambridge, M.A.

No context found.

Jorgen W. Weibull. Evolutionary Game Theory. MIT Press, Cambridge, Massachusetts, 1995.

No context found.

J. W. Weibull. Evolutionary Game Theory. MIT Press, 1995.

No context found.

Jorgen Weibull. Evolutionary Game Theory. MIT Press, Cambridge, MA, 1995.

No context found.

Weibull, J.W., Evolutionary Game Theory, MIT Press, (1996).

No context found.

J. W. Weibull. Evolutionary Game Theory. MIT Press, 1995.

No context found.

J. W. Weibull. Evolutionary Game Theory. The MIT Press, 1995.

No context found.

J. W. Weibull. Evolutionary Game Theory. MIT Press, Cambridge, MA, 1995.

No context found.

J. W. Weibull, Evolutionary Game Theory , MIT Press, Cambridge, MA (

No context found.

Jorgen W. Weibull. Evolutionary Game Theory. MIT Press, 1995.

No context found.

J orgen W. Weibull. Evolutionary Game Theory. The MIT Press, Cambridge, MA, 1995.

No context found.

Weibull, J.W., (1995), Evolutionary Game Theory, Cambridge, Mass., The MIT Press.

No context found.

Weibull, J. W. (1995), Evolutionary Game Theory, MIT Press, Cambridge.

No context found.

J. W. Weibull, Evolutionary Game Theory, The MIT Press, 1995.

No context found.

Weibull, J.(1995), "Evolutionary Game Theory", MIT Press.

No context found.

Weibull, J. W. (1995), Evolutionary Game Theory, MIT Press, Cambridge.

No context found.

J. Weibull. Evolutionary Game Theory. MIT Press, 1996.

No context found.

Weibull, J. W. (1995), Evolutionary Game Theory, MIT Press, Cambridge.

No context found.

Weibull, J.W. (1995) Evolutionary Game Theory, MIT Press, Cam- bridge.

First 50 documents  Next 50

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC