聚类算法 Kmeans

KMeans 聚类算法

    伪代码:
	seeds = random(data , k)
	#seeds cluster center 
	#k  cluster counts
	repeat:
		for data in datas;    
			find data min distance in seeds and update data cluster = k 
		update cluster center 
      	until cluster doesn't change(center change rate small than you set diff quota)
	

Kmeans难点

KMeans家族

参考

Fork me on GitHub