Paper Info
Title
Choosing the kernel parameters for support vector machines by the inter-cluster distance in the feature space
Abstract
Determining the kernel and error penalty parameters for support vector machines (SVMs) is very problem-dependent in practice. A popular method to deciding the kernel parameters is the grid search method. In the training process, classifiers are trained with different kernel parameters, and only one of the classifiers is required for the testing process. This makes the training process time-consuming. In this paper we propose using the inter-cluster distances in the feature spaces to choose the kernel parameters. Calculating such distance costs much less computation time than training the corresponding SVM classifiers; thus the proper kernel parameters can be chosen much faster. Experiment results show that the inter-cluster distance can choose proper kernel parameters with which the testing accuracy of trained SVMs is competitive to the standard ones, and the training time can be significantly shortened.
Year
DOI
Venue
2009
10.1016/j.patcog.2008.08.030
Pattern Recognition
Keywords
Field
DocType
support vector machine,distance cost,support vector machines,training time,feature space,training process,inter-cluster distance,different kernel parameter,computation time,kernel parameter,svm support vector machines kernel parameters inter-cluster distances,kernel parameters,testing process,grid search method,inter-cluster distances,proper kernel parameter,svm
Graph kernel,Pattern recognition,Radial basis function kernel,Kernel embedding of distributions,Tree kernel,Polynomial kernel,Artificial intelligence,Kernel method,String kernel,Variable kernel density estimation,Mathematics,Machine learning
Journal
Volume
Issue
ISSN
42
5
Pattern Recognition
Citations 
PageRank 
References 
49
1.52
11
Authors
2
Name
Order
Citations
PageRank
Kuo-Ping Wu123712.06
Sheng-De Wang272068.13