Special Issue from the International Conference on Swarm Intelligence: Theoretical Advances and Real-World Applications-June 14th and 15th,2011 at the Ecole Internationale des Sciences du Traitement de I'Information (EISTI), Cergy, France

摘要

There is no simple and comprehensive definition of "swarm intelligence," and that is the way it should be. Generally speaking, in computational intelligence research a swarm comprises numerous entities, where knowledge, problem solving, or search is distributed across them and individuals persist over iterations. Though there have been attempts to eliminate randomness from some of the algorithms, most swarm implementations are stochastic to some degree. Swarms are unpredictable, with some kind of shaky truce negotiated between coordination and disorder. The field is often thought to be dichotomized between particle swarms and ant algorithms, but there are many extensions, adaptations, novelties, and innovations, and considerable overlap with evolutionary methods; there is little resistance to sharing techniques across research stovepipes, and the field of swarm intelligence itself can be visualized as a kind of metaswarm of researchers and methods.