Efficient Approximation for Large-Scale Kernel Clustering Analysis

Kernel k-means is useful for performing clustering on nonlinearly separable data. The kernel k-means is hard to scale to large data due to the quadratic complexity. In this paper, we propose an approach which utilizes the low-dimensional feature approximation of the Gaussian kernel function to capitalize a fast linear k-means solver to perform the nonlinear kernel k-means. This approach takes advantage of the efficiency of the linear solver and the nonlinear partitioning ability of the kernel clustering. The experimental results show that the proposed approach is much more efficient than a normal kernel kmeans solver and achieves similar clustering performance.