A Novel Evolutionary Programming in Clustering of Multi-Dimensional Vector Space

Abstract

Recently, Evolutionary Computation (EC) schemes have been implemented into the clustering of the vector space. However, the existing schemes face to some problems on finding the optimal solution. There are two major difficulties in the traditional system. It spends too much time for the system to find a nearly optimal solution. Also, the aspect toward the global solution at the end of the process would be unclear. In this paper, a novel EC scheme called ''Sub-Individual Evolutionary Programming'' (SIEP) is suggested to suppressed the problems that is found in the conventional system. From the experimental result, the number of generation for approaching optimal solution of the new scheme is notably smaller than the existing schemes.