Anchor graph semi-supervised learning method based on Evidence Theory

In many practical machine leaning tasks, labeled training samples are expensive to obtain, while unlabeled ones are readily available. Therefore, semi-supervised classification(SSC) have attracted a lot of attention, among which graph based semi-supervised learning is one of the main research hotspots. The Graph semi-supervised learning methods use unlabeled samples and few labeled samples to build a reliable graph model and spread the label information on the graph model. However, the traditional graph semi-supervised learning algorithms are time-consuming and can not build very reliable graph models on high-dimensional data. In order to deal with these problems, this paper propose an anchor graph semi-supervised learning method based on Evidence Theory. The proposed method uses some multiple feature subspaces of original data to train different anchor graph semi-supervised learning models, and combines the different results using Evidence Theory. Experimental results based on benchmark data sets show that the proposed algorithm can achieve better performance compared with the prevailing related algorithms.