Research of K-means clustering method based on DNA genetic algorithm and P system

Abstract

This paper proposed a k-means clustering analysis method, which is based on DNA genetic algorithm and P system. DNA encoding is used to analyze the initial center of cluster and the P system is used to realize the clustering. The quality of the clustering is judged by the Euclidean distance of the corresponding cluster center, the rate of convergence and the image of the clustering result. Through selection, crossover, mutation and inversion, we can get the best center of the cluster. At the end of the paper, we take the simulation experience and the result shows the effect of this method is superior to the genetic algorithm of k-means clustering method. The simulation experiment is finished in the MATLAB 2014a and the experiment data is random generated.