改进的k-means方法，对聚类的实例节能型加权 少数类多数类的函数-Improved k-means clustering instance, energy-efficient weighted minority class and majority class of functions

针对于高维稀疏的子空间数据，并进行了加权熵处理(Aiming at the high-dimensional sparse subspace data, weighted entropy processing is carried out.)