The research of the parallel SMO algorithm for solving SVM

Author

Peng, Peng ; Ma, Qian-li ; Hong, Lei-ming

Author_Institution

South China Univ. of Technol., Guangzhou, China

Volume

3

fYear

2009

fDate

12-15 July 2009

Firstpage

1271

Lastpage

1274

Abstract

In order to improve solving support vector machine algorithm, an improved learning algorithm of the parallel SMO is proposed. According to this algorithm, the master CPU averagely distributes primitive training set to slave CPUs so that they can almost independently run serial SMO on their respective training set. As it adopts the strategies of buffer and shrink, the speed of the parallel training algorithm is increased, which is showed in the experiments of parallel SMO based on the dataset of MNIST. The experiments indicate that the parallel SMO algorithm has good performance in solving largescale SVM.

Keywords

algorithm theory; learning (artificial intelligence); minimisation; support vector machines; buffer; learning algorithm; master CPU; parallel SMO algorithm; primitive training set; sequential minimal optimisation; serial SMO; slave CPU; support vector machine algorithm; Cybernetics; Kernel; Large-scale systems; Machine learning; Machine learning algorithms; Master-slave; Pattern recognition; Probability density function; Support vector machine classification; Support vector machines; Learning algorithm; Parallel SMO; Support Vector Machine;

fLanguage

English

Publisher

ieee

Conference_Titel

Machine Learning and Cybernetics, 2009 International Conference on

Conference_Location

Baoding

Print_ISBN

978-1-4244-3702-3

Electronic_ISBN

978-1-4244-3703-0

Type

conf

DOI

10.1109/ICMLC.2009.5212348

Filename

5212348