A Fast and Accurate Progressive Algorithm for Training Transductive SVMs

Authors:
Lei Wang;Huading Jia;Shixin Sun
Affiliations:
School of Computer Science & Engineering, University of Electronic Science & Technology of China, Chengdu, 610054, P.R. China;Institute of Image & Graphics, Sichuan University, Chengdu, 610064, P.R. China and School of Economics Information Engineering, Southwest University of Finance & Economics, Chengdu, 610074, P.R. C ...;School of Computer Science & Engineering, University of Electronic Science & Technology of China, Chengdu, 610054, P.R. China
Venue:
ISNN '07 Proceedings of the 4th international symposium on Neural Networks: Advances in Neural Networks, Part III
Year:
2007

Citing 7
Cited 0

The nature of statistical learning theory

The nature of statistical learning theory
Bagging predictors

Machine Learning
Making large-scale support vector machine learning practical

Advances in kernel methods
Transductive Inference for Text Classification using Support Vector Machines

ICML '99 Proceedings of the Sixteenth International Conference on Machine Learning
Training TSVM with the proper number of positive samples

Pattern Recognition Letters
Deterministic annealing for semi-supervised kernel machines

ICML '06 Proceedings of the 23rd international conference on Machine learning
Confidence-based classifier design

Pattern Recognition

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper develops a fast and accurate algorithm for training transductive SVMs classifiers, which utilizes the classification information of unlabeled data in a progressive way. For improving the generalization accuracy further, we employ three important criteria to enhance the algorithm, i.e. confidence evaluation, suppression of labeled data, stopping with stabilization. Experimental results on several real world datasets confirm the effectiveness of these criteria and show that the new algorithm can reach to comparable accuracy as several state-of-the-art approaches for training transductive SVMs in much less training time.