Introduction - If you have any usage issues, please Google them yourself
Algorithm Description: K-means clustering algorithm: Given the number of classes K, the N objects assigned to a class K to go, making within-class similarity between objects the most, while the similarity between the smallest class