An improved incremental training algorithm for support vector machines using active query
Document Type
Article
Publication Date
3-1-2007
Abstract
In this paper, we present an improved incremental training algorithm for support vector machines (SVMs). Instead of selecting training samples randomly, we divide them into groups and apply the k-means clustering algorithm to collect the initial set of training samples. In active query, we assign a weight to each sample according to its confidence factor and its distance to the separating hyperplane. The confidence factor is calculated from the error upper bound of the SVM to indicate the closeness of the current hyperplane to the optimal hyperplane. A criterion is developed to eliminate non-informative training samples incrementally. Experimental results show our algorithm works successfully on artificial and real data, and is superior to the existing methods. © 2006 Pattern Recognition Society.
Identifier
33750506073 (Scopus)
Publication Title
Pattern Recognition
External Full Text Location
https://doi.org/10.1016/j.patcog.2006.06.016
ISSN
00313203
First Page
964
Last Page
971
Issue
3
Volume
40
Fund Ref
National Science Foundation
Recommended Citation
Cheng, Shouxian and Shih, Frank Y., "An improved incremental training algorithm for support vector machines using active query" (2007). Faculty Publications. 13507.
https://digitalcommons.njit.edu/fac_pubs/13507
