一种基于支持向量机的直推式学习算法

A Transductive Learning Algorithm Based on Support Vector Machine

直推式支持向量机(Transductive Support Vector Machine,TSVM)是标准的支持向量机算法在半监督学习问题上的一种扩展,但已有的TSVM算法存在训练速度慢、回溯式学习多、学习性能不稳定等缺点,针对这些问题提出一种改进的直推式支持向量机算法———ITSVM,该算法较准确地确定了待训练的未标识样本中的正负样本数问题,有效解决了传统TSVM中过多的回溯式学习问题,同时该算法也无需利用过多的未标识训练样本,从而减轻了计算强度.实验表明,ITSVM相比TSVM在分类正确率、分类速度以及使用的样本规模上,都表现出了一定的优越性.

Transductive Support Vector Machine （TSVM） is an extension of standard SVM in semi-supervised learning problem. However, the existing TSVM algorithm has some drawbacks such as slower training speed, more back learning steps, and unstable learning performance, etc. This paper presents an improved transductive Support Vector Machine learning algorithm——ITSVM, which can determine the positive labeled sample numbers of the up-training unlabeled samples, and alleviate the back learning phenomenon. Besides more unlabeled training samples are not needed, and where by the calculation complexity can be reduced. The experiment results show that ITSVM is better than TSVM in speed and accuracy of classification.