摘要
通过比较目标函数、聚类原型模式P(0)的初始化方法、划分矩阵U和聚类原型P的更新方法等4个方面,得出k均值算法和硬C-均值算法的区别。
Differences between k-means algorithm and HCM algorithm have been obtained through contrasting objective functions,initializing cluster prototype pattern P(0),dividing matrix U and upgrading cluster prototype P.
出处
《洛阳理工学院学报(自然科学版)》
2012年第1期72-75,81,共5页
Journal of Luoyang Institute of Science and Technology:Natural Science Edition