Study on a New RPCCL Clustering Algorithm
Sanfeng Chen, Tao Mei, Minzhou Luo, Huawei Liang · 2007
Based on further study on the existing RPCCL (rival penalized controlled competitive learning) algorithms and discovering some drawbacks, a new advanced RPCCL clustering algorithm is proposed. The new algorithm initializes the clustering centers according to the density of sample points, and the weights are updated according to the location of input data. The simulation results show that the new advanced algorithm can perform clustering more accurately and rapidly.