% ihlf_sv_rfn toolkit
% version 1.0
%
% abaloneexample.m - demonstrate the recusive finite Newton algorithm on Abalone data set
% evalkernel.m - evaluate the kernel function
% ihlf_svr_rfntrain.m - interface of the following three types of algorithms
% ihlf_svr_rfntrain_fast.m - precompute the entire kernel matrix and space complexity is O(n*n)
% ihlf_svr_rfntrain_normal.m - precompute the n*nsv kernel matrix  and space complexity is O(n*nsv)
% ihlf_svr_rfntrain_save.m - do not precompute the kernel matrix and space complexity is O(nsv*nsv)
% ihlf_svr_rfntest.m - evaluate the estimated output of the input
% scaletrain.m - normalize data in the interval [-1 1]
% scaletest.m - normalize data in the interval [-1 1]