数据挖掘中Fuzzy c-means的自适应聚类算法

A self-adaptive algorithm based on fuzzy c-means in data mining

聚类算法是数据挖掘算法中的重要解决方法。针对现有聚类算法模糊c均值算法FCM中的不足 ,如需要预先确定聚类参数c,随机性较强、局部最优等弱点 ,对其算法结构加以改进 ,提出模糊c均值自适应算法 (FCMA) ,增加聚类有效性问题的分析 ,在聚类过程中可动态调整聚类数目 ,这种方法可以避免在确定参数时的随机性和经验性 ,提高聚类算法的可靠程度。

Clustering algorithm is an important one in data mining methods. In this article, the self adaptive algorithm based on fuzzy c means(fcm) was broughtout, which can derivation the clustered number automatically by a validity function. This algorithm can avoid the randomicity and experientialism in the clustering process , so it is effective.