Random feature based multiple kernel clustering
Jin Zhou, Yuqi Pan, Lin Wang, C. L. Philip Chen · 2016
The kernel clustering method is very helpful in non-linear data clustering. But its high computational complexity makes it unattainable to large datasets. In this paper, a new multi-kernel clustering algorithm based on the random Fourier feature is proposed to solve this issue, where the maximum-entropy method is applied to optimize the kernel weights. Experiment on synthetic non-linear dataset has shown the good performance of the proposed algorithm.