作者: Jansen , Wegener
DOI: 10.1007/S00453-002-0940-2
关键词:
摘要: Evolutionary algorithms are randomized search heuristics that were invented in the sixties and have been intensively applied studied since eighties. Since then there only a few theoretical investigations no sound foundation. One of main sources difficulty for analyses is crossover operator. It can be useful if current population strings has certain diversity. Here it proved an evolutionary algorithm produce enough diversity such use speedup expected optimization time from superpolynomial to polynomial small degree.