Title :
Coevolutionary learning in the tragedy of the commons
Author :
Garcia, Julián ; Nino, F.
Author_Institution :
Dept. of Comput. & Syst. Eng., National Univ. of Colombia, Bogota, Colombia
Abstract :
The joint utilisation of a commonly owned resource often causes the resource to be overused. This is known as the tragedy of the commons. This paper analyses the effects of coevolutionary learning in such kind of situations using genetic programming. In a game-theoretical approach, the situation considers not only the strategic interaction among players, but also the dynamics of a changing environment linked strongly to the players´ actions and payoffs. The results of an analytical game are used to formulate a simulation game for the commons, then a series of computational experiments are conducted, obtaining coevolved game strategies that are examined in comparison with those predicted by the analytical model. The obtained results are similar to those predicted by classic game theory, but not always leading to a tragedy.
Keywords :
evolutionary computation; game theory; learning (artificial intelligence); coevolutionary learning; game theory; genetic programming; simulation game; tragedy-of-the-commons; Analytical models; Aquaculture; Artificial intelligence; Computational modeling; Game theory; Genetic programming; Global warming; Learning; Predictive models; Systems engineering and theory;
Conference_Titel :
Evolutionary Computation, 2003. CEC '03. The 2003 Congress on
Print_ISBN :
0-7803-7804-0
DOI :
10.1109/CEC.2003.1299945