摘要
在各种糊聚类算法中 ,模糊C -均值聚类算法FCM (FuzzyC MeanClusteringAlgorithm )的应用最为广泛。但在实际的应用中 ,FCM算法却容易陷入局部最优解。因此 ,本文首先提出了一种基于遗传算法GA(GeneticAlgorithm )的模糊聚类分析方法 ,它利用了遗传算法随机搜索的特点 ,可以避免陷入局部最优解。实验表明 ,将该遗传算法与FCM算法结合起来进行聚类分析 ,比单一使用遗传算法或单一使用FCM算法进行聚类分析的效果都要好。
Among all the clustering algorithms, Fuzzy C Mean clustering algorithm (FCM) is widely used. But in practical applications, the FCM easily plunges into the local optimum. This paper proposes a fuzzy clustering algorithm based on Genetic Algorithm (GA), which can avoid plunging into the local optimum, because it searches the solution randomly. But using it can only find the solution close to the global optimum, and can't absolutely assure that it converges on the global optimum. Then this paper proposes anther algorithm that integrates the GA with the FCM.The examples show that the results are better than those of only using the GA or the FCM.
出处
《华东船舶工业学院学报》
EI
2001年第6期40-43,共4页
Journal of East China Shipbuilding Institute(Natural Science Edition)
