1 Introduction
Projection pursuit can convert high-dimensional data
into low-dimensional spaces. The characteristics of the
high-dimensional data are studied by analyzing the
projection features of data of low-dimensional spaces.
A Projection Pursuit Cluster (PPC) model is a
method for dealing with complex multi-factor statistical
problems[1, 2]. In this method, different projection
directions reflect different characteristics of the data
structure. The main difficulties of establishing this