不规则格阵的Potts模型在数据聚类(分团)中的应用

Application of Irregular Potts Model in Data Clustering

介绍了一种新式的聚类方法。它将每个数据点都赋予一个Potts自旋值,并引入一个相邻数据点间的短程作用力。当体系处于超顺磁相时,会出现一些由自旋子构成的自组织区域,我们利用蒙特卡洛方法来测量此时自旋子间的相关系数,用来将数据点进行聚类。在整个问题的解决过程中,对数据不作任何的假设。

The present paper introduces a new clustering technique. A Potts spin is assigned to each data point and short-range interactions between neighboring points are introduced. Spin-spin correlations, measured （by Monte Carlo） in a super-paramagnetic regime in which aligned self-organization appears, serve to partition the data points into clusters. In the whole process, we do not assume any structure of the data, just based on the underlying distribution of it.