One research of clustering algorithm based on rough set and genetic algorithm

摘要

This paper integrates rough set theory and adaptive genetic algorithm to propose a new clustering method based on symbols attributes (RAGACA). For each different value, the algorithm adopts the top-down divisive hierarchical clustering strategy and uses RAGA algorithm to dichotomize the data sets step by step until it reaches pre-specified number of cluster, and then output the clustering result. The experimental results show that RACACA has higher accuracy and convergence for the data with symbols attributes.