Efficient Kernel Approximation for Large-Scale Support Vector Machine Classification

Training support vector machines (SVMs) with nonlinear kernel functions on large-scale data are usually very timeconsuming. In contrast, there exist faster solvers to train the linear SVM. We propose a technique which sufficiently approximates the infinite-dimensional implicit feature mapping of the Gaussian kernel function by a low-dimensional feature mapping. By explicitly mapping data to the low-dimensional features, efficient linear SVM solvers can be applied to train the Gaussian kernel SVM, which leverages the efficiency of linear SVM solvers to train a nonlinear SVM. Experimental results show that the proposed technique is very efficient and achieves comparable classification accuracy to a normal nonlinear SVM solver.