作者: Xingong Chang
关键词:
摘要: A cooperative coevolutionary EA based algorithm is developed to discover potentially useful substructures from graphical databases. Unlike the usual algorithms which are on divide-and-conquer strategy with different populations representing subtasks, cooperation in our at individual-level and implemented by a new genetic operator, individual operator. The during searching process, enables individuals search same substructure way hence handles problem of losing instances, very common vital performance. In addition, an approximate graph matching also proposed make operator more efficient. Experimental results show that successfully enhances capability improves qualities solutions.