Cultural Algorithm文化算法.ppt

CULTURAL SWARMS: KNOWLEDGE-DRIVEN PROBLEM SOLVING IN SOCIAL SYSTEMS Robert G. Reynolds The Emergence of Social Intelligence and Socially Motivated Problem Solving Traditionally Artificial Intelligence has focused on the nature of “individual intelligence”. Goal was to create programs that compete with the best “individual”. E.G. Deep Blue in chess. With the advent of high performance distributed networks of computers and grid computation there is much incentive to investigate socially motivated problem solving. Example frameworks have come from naturally occurring systems. The Emergence of Social Intelligence and Socially Motivated Problem Solving Continued. Particle swarms (colloids) from physical systems. Ant Colony Optimization from Biological Systems. Cultural Algorithms from Cultural Systems. Allows us to begin to understand the nature of social intelligence in terms of group problem solving. Socially Motivated Approaches Particle Swarm Optimization (Kennedy and Eberhart 1995) Based on social behaviors of birds and fish Particles (potential solutions) move by following its own best and best of its neighbor Ant Colony Optimization (Dorigo et. al. 1996 ) Based on ant colony observation the pheromone and the density of the pheromone along the trail is the knowledge that the ACO shares among its individual ants Cultural Algorithms as a Vehicle for Socially Motivated Problem Solving Cultural Algorithms consist of a social population and a belief space [Reynolds, 1979] Derived from models of Cultural Evolution in Anthropology [Flannery,2000, 1986, 1969]. Provides a framework to accumulate and communicate knowledge so as to allow self-adaptation in an evolving model. Encompasses the scales of the other approaches so should be able to solve their types of problems as well. Cultural Algorithms Have been applied to modeling many problems in the evolution of complex social organizations. E.G. Simulation of the Ancient Anazazi (Kohler, Gummerman, and Reynolds, Sci