Title :
Multi-scale Least Square Wavelet Support Vector Machine
Author :
Wang Xianfang ; Wu Ruihong ; Cui Jinling
Author_Institution :
Sch. of Comput. & Inf. Technol., Henan Normal Univ., Xinxiang, China
Abstract :
Original Least square Wavelet Support Vector Machine (LSSVM) algorithm can not reach desired precision in multi-scale regression. To solve the problem, a multi-scale wavelet LSSVM algorithm is proposed in this paper by using a wavelet kernel. Mexican-hat wavelet function is used as the support vector kernel function, and further the Least square Wavelet Support Vector Machine (LS-WSVM) algorithm is presented. On this basis, the global optimum of the multi-scale regression modeling problem can be obtained by solving a quadratic programming problem. As a result, the regression model can effectively approximate multi-scale signals. Therefore, LS-WSVM is an efficient modeling method and worth popularization and application by computer simulation results.
Keywords :
approximation theory; least squares approximations; quadratic programming; regression analysis; support vector machines; LS-WSVM algorithm; LSSVM algorithm; Mexican-hat wavelet function; computer simulation; multiscale least square wavelet support vector machine; multiscale regression modeling problem; multiscale signal approximation; quadratic programming problem; support vector kernel function; wavelet kernel; Approximation algorithms; Function approximation; Kernel; Least squares approximation; Mathematical model; Support vector machines; Least square support vector machine; Multi-scale regression; Wavelet;
Conference_Titel :
Control and Decision Conference (CCDC), 2012 24th Chinese
Conference_Location :
Taiyuan
Print_ISBN :
978-1-4577-2073-4
DOI :
10.1109/CCDC.2012.6243066
