多尺度变换近邻传播算法

MDS-based affinity propagation clustering algorithm

利用多尺度变换将高维数据映射成低维数据,便于近邻传播聚类。仿真证明,与传统近邻传播算法相比,基于多尺度变换的近邻传播算法聚类精度高,收敛速度快。

By means of multidimensional scaling （MDS） ,the high dimensional data is mapped into the low dimensional one for convenient affinity propagation clustering . Simulations show that the method ,compared with the traditional affinity propagation , is with higher precision and higher convergence speed .