Introduction - If you have any usage issues, please Google them yourself
The between-cluster variance as the basic idea is to use a threshold to the whole data is divided into two classes, if the maximum variance between two classes, then the threshold value is the best threshold value