Fast forecasting with simplified kernel regression machines

Kernel machines, including support vector machines, regularized networks and Gaussian process etc, have been widely used in forecasting. However standard algorithms are often time consuming. To this end, we propose a new method for imposing the sparsity of kernel regression machines. Different to previous methods, it incrementally finds a set of basis functions that minimizes the primal cost functions directly. The main advantage of out method lies in its ability to form very good approximations for kernel regression machines with a clear control on the computation complexity as well as the training time. Experiments on two real time series and benchmark Sunspot assess the feasibility of out method.