Comparative Studies of Two Algorithmsfor Solving Single-output Nonlinear LS-SVR
HOU Min
ZHAO Chun-hui
Abstract：For solving single‐output nonlinear least squares support vector regression machine(NLS‐SVR) modeling ,scholars presented two methods ,classical algorithm(C‐NLS‐SVR) and fast algorithm (F‐NLS‐SVR) .In order to examine their advantages and disadvantages ,this paper provides detailed theoretical comparison and performs a series of comparative experiments on 10 datasets taken from UCI database and 4 evaluating indicators .Experiment results indicates that the running time speeded upby F‐NLS‐SVR is not much ,but the loss of the fitting accuracy is relatively large .
Keywords：least squares support vector regression machineclassical algorithmfast algorithmal-gorithm comparisonkernel function
Publication Date：2016-01-01
Online Publishing Date：2026-08-28(First online date of this platform, not the publication date of the document)
Pages：7( 8-13,37 )