Title :
A hybrid decomposition/interior point algorithm for massive support vector machine
Author :
Li, Jin ; Liu, Jing-Xu ; Tan, Yue-jin ; Liao, Liang-Cai
Author_Institution :
Dept. of Manage., Nat. Univ. of Defense Technol., Changsha, China
Abstract :
A hybrid decomposition/interior point algorithm for training SVM has been proposed. The hybrid method applies an interior point method to the quadratic programs arising from decomposition algorithm, and then some techniques are presented to improve the performance of algorithm. It has been demonstrated that the algorithm is applicable for training SVM, especially effective for massive SVM.
Keywords :
learning (artificial intelligence); quadratic programming; support vector machines; SVM training; hybrid decomposition-interior point algorithm; quadratic program; support vector machine; Cybernetics; Equations; Large-scale systems; Machine learning; Management training; Neural networks; Quadratic programming; Support vector machine classification; Support vector machines; Technology management; Decomposition; Hybrid; Interior Point; Support Vector Machine;
Conference_Titel :
Machine Learning and Cybernetics, 2005. Proceedings of 2005 International Conference on
Conference_Location :
Guangzhou, China
Print_ISBN :
0-7803-9091-1
DOI :
10.1109/ICMLC.2005.1527689
