DocumentCode
532960
Title
A parallel approach for SVM with multi-core CPU
Author
Hu, Mingsheng ; Hao, Weixu
Author_Institution
Inst. of Software Sci., Zhengzhou Normal Univ., Zhengzhou, China
Volume
15
fYear
2010
fDate
22-24 Oct. 2010
Abstract
Modern multi-core CPU are predominant in improving performance of parallel applications. We describe an approach for support vector machines (SVM) that can be parallelized efficiently and scales to large problems with large training vectors. Parallel programming in .NET framework using SVM on multi-core computers are studied. Experiment results show that the serial program can only use one core of multi-core computers. Meanwhile, the parallel efficiency of the program using SVM is very high on 2-core and 4-core computers respectively.
Keywords
multiprocessing systems; parallel programming; support vector machines; .NET framework; SVM; multicore CPU; parallel programming; support vector machine; Classification algorithms; Computational modeling; Educational institutions; Programming; Support vector machines; Parallel programming; SVM; multi-core;
fLanguage
English
Publisher
ieee
Conference_Titel
Computer Application and System Modeling (ICCASM), 2010 International Conference on
Conference_Location
Taiyuan
Print_ISBN
978-1-4244-7235-2
Electronic_ISBN
978-1-4244-7237-6
Type
conf
DOI
10.1109/ICCASM.2010.5622595
Filename
5622595
Link To Document