Title :
Selection of support vector machines parameters for regression using nested grids
Author :
Popov, Alexander ; Sautin, Alexander
Author_Institution :
Dept. of Software & Database Eng., NSTU, Novosibirsk
Abstract :
The paper examines support vector machines for regression problem. Analysis of different grid types for selection of SVM parameters is conducted. Experimental results obtained on the basis of applying nested grids are presented, and efficiency of regression problem solution with various support vector machine parameters values is investigated.
Keywords :
regression analysis; support vector machines; SVM parameter selection; nested grids; regression problem; support vector machines; Constraint optimization; Data engineering; Databases; Equations; Information technology; Kernel; Machine learning; Search methods; Support vector machines; Training data;
Conference_Titel :
Strategic Technologies, 2008. IFOST 2008. Third International Forum on
Conference_Location :
Novosibirsk-Tomsk
Print_ISBN :
978-1-4244-2319-4
Electronic_ISBN :
978-1-4244-2320-0
DOI :
10.1109/IFOST.2008.4602974
