A partitioning method of multicenters
Ning Yonghen · 2014
A multicenter partitioning clustering algorithm(DMC)was proposed.In DMC,the single center was substituted with the multicenter.The shortest center was defined as the partition goal,and the cluster result was obtained by adjusting the multicenter.Experimental results showed that DMC could overcome the shortcomings of the tendency to identify similar sizes and the convex distribution of data in the partitioning method.It can solve the problem of different data distribution and can obtain a better result without increasing complexity.