A Novel Hybrid High-Dimensional PSO Clustering Algorithm Based on the Cloud Model and Entropy

Abstract

With the increase in the number of high-dimensional data, the characteristic phenomenon of unbalanced distribution is increasingly presented in various big data applications. At the same time, most of the existing clustering and feature selection algorithms are based on maximizing the clustering accuracy. In addition, the hybrid approach can effectively solve the clustering problem of unbalanced data. Aiming at the shortcomings of the unbalanced data clustering algorithm, a hybrid high-dimensional multi-objective PSO clustering algorithm is proposed based on the cloud model and entropy (HHCE-MOPSO). Furthermore, the feasibility of the hybrid PSO is verified by the simulation of the multi-objective test function. The results not only broaden the new theory and method of clustering algorithm for unbalanced data, but also verify the accuracy and feasibility of the hybrid PSO. Furthermore, the clustering analysis method based on information entropy is a new method. As a result, the research results have both important scientific value and good practical significance.