具有短暂记忆的狼群搜索算法 WSA（Wolf search algorithm with ephemeral memory）是一种新的群智能优化算法，然而该算法存在易陷入局部最优之不足。为此，提出一种改进的狼群搜索算法 IWSA（An improved wolf search algorithm）。将该算法应用于解决聚类问题，提出一种 IWSA 与 k-means 相结合的聚类方法。方法在每次迭代中，利用 IWSA 来指导搜寻全局最优质心，避免了算法陷入局部最优。最后，用11个真实数据集和2个人工合成数据集来测试该方法的性能。测试结果证实了该方法是可行的和有效的。%Wolf group search algorithm with ephemeral memory (WSA)is a new swarm intelligence optimization algorithm,but it has the drawback of easy falling into local optima.In order to overcome the deficiency of the WSA,an improved wolf group search algorithm with e-phemeral memory (IWSA)is proposed in this paper.Besides,this IWSA is applied to solve the clustering problem,and a new clustering meth-od called clustering method which combines IWSA with k-means is presented in the paper.This method uses the IWSA to guide to search for global optimal centroid in each iterative computation,avoiding local optima.We use eleven real datasets and two synthetic datasets to test the performance of the proposed method,and the test results validate that the proposed method is feasible and effective.
展开▼
