A Novel Business Administration Big Data Clustering Optimization Method Based On Active Density Peak And Salpa Swarm Algorithm

Big data clustering plays an important role in the field of business administration data processing, but the big data clustering method has some problems such as poor clustering effect and low Jaccard coefficient. Therefore, this paper proposes a novel business administration big data clustering optimization method based on active density peak and Salpa swarm algorithm. This method combines principal component analysis method and information entropy dimensionality reduction to process big data, reduces the time required for data clustering, and uses intuitionistic fuzzy kernel clustering algorithm to cluster big data. The algorithm uses the Logic-Tent chaotic sequence to promote the uniform distribution of the initial population of Salp swarm algorithm (SSA). By adding attenuation factors that limit the leader's over-searching, the global exploration ability and local exploitation ability of the algorithm are balanced. Adaptive inertia weights are added to follower updates to avoid convergence of the algorithm to local extreme values. This method designs a fast update strategy and effectively updates labels. An active cluster integration framework combining local and global uncertainties is proposed, and weighted voting consistency method is introduced to optimize the integration of cluster results. Experimental results on data sets show that the proposed method can effectively solve the problems of traditional methods and has good clustering performance.