Abstract：For the fuzzy clustering C-means algorithm is sensitive to the initial clustering center and clustering slow convergence, manually set the number of clusters and other defects, based on the density peak value, we measure the density of data and the distance between cluster centers, so as to automatically determine the number of cluster centers and clusters, which is used as the initial center of the improved bat algorithm. In this paper, the Levy flight characteristics are introduced to enhance the bat algorithm to jump out of the local optimum ability, and Powell local search is used to accelerate bat algorithm convergence.In this paper, the improved bat population is used for population optimization, and the optimal bat position is used as the clustering C-means new clustering center, and fuzzy clustering is carried out to obtain the clustering results by repeated iterative iterations.Compared with the other two clustering algorithms on the standard dataset, the experimental results show that compared with the other two algorithms, the clustering algorithm proposed in this paper can reduce the number of iterations and converge quickly, lower error rate.
2020, Vol. 41 
