Reddy, Sathish I and Shevade, Shirish and Murty, MN (2011) A fast quasi-Newton method for semi-supervised SVM. In: Pattern Recognition, 44 (10-11,). pp. 2305-2313.
A_fastquasi-.pdf - Published Version
Restricted to Registered users only
Download (479Kb) | Request a copy
Due to its wide applicability, semi-supervised learning is an attractive method for using unlabeled data in classification. In this work, we present a semi-supervised support vector classifier that is designed using quasi-Newton method for nonsmooth convex functions. The proposed algorithm is suitable in dealing with very large number of examples and features. Numerical experiments on various benchmark datasets showed that the proposed algorithm is fast and gives improved generalization performance over the existing methods. Further, a non-linear semi-supervised SVM has been proposed based on a multiple label switching scheme. This non-linear semi-supervised SVM is found to converge faster and it is found to improve generalization performance on several benchmark datasets. (C) 2010 Elsevier Ltd. All rights reserved.
|Item Type:||Journal Article|
|Additional Information:||Copyright of this article belongs to Elsevier Science.|
|Keywords:||Semi-supervised learning;Support vector machines;Quasi-Newton methods;Nonconvex optimization|
|Department/Centre:||Division of Electrical Sciences > Computer Science & Automation (Formerly, School of Automation)|
|Date Deposited:||05 Aug 2011 09:07|
|Last Modified:||05 Aug 2011 09:07|
Actions (login required)