关键词: 大规模数据集谱聚类, 加权核K-means算法, 数据抽样, 核矩阵

Abstract:

Classical spectral clustering algorithm transforms data clustering into graph partitioning problems, so, based on analyzing the equivalence between its Normalized Cut objective function and the weighted nuclear K-means function, a large-scale data spectrum based on sampling improved weighted nuclear K-means is designed, in which initial clustering preprocessing by Leaders is used to control the size of subsequent random sampling data and coverage of the original data categories, and the weighted kernel K-means iterative optimization is used to avoid the large resource consumption of Laplacian matrix feature decomposition of classical spectral clustering algorithm, thereby avoiding the time-space complexity of all nuclear moments by using of partial kernel matrices. Experimental results show that, the improved algorithm can greatly improve the clustering efficiency on the basis of maintaining similar clustering accuracy with the classic algorithm.

Key words: big data spectral clustering, weighted kernel K-means, data sampling, kernel matrix