Estimate Unlabeled-Data-Distribution for Semi-supervised PU Learning

Abstract

Traditional supervised classifiers use only labeled data (features/label pairs) as the training set, while the unlabeled data is used as the testing set. In practice, it is often the case that the labeled data is hard to obtain and the unlabeled data contains the instances that belong to the predefined class beyond the labeled data categories. This problem has been widely studied in recent years and the semi-supervised learning is an efficient solution to learn from positive and unlabeled examples(or PU learning). Among all the semi-supervised PU learning methods, it's hard to choose just one approach to fit all unlabeled data distribution. This paper proposes an automatic KL-divergence based semi-supervised learning method by using unlabeled data distribution knowledge. Meanwhile, a new framework is designed to integrate different semi-supervised PU learning algorithms in order to take advantage of the former methods. The experimental results show that (1)data distribution information is very helpful for the semi-supervised PU learning method; (2)the proposed framework can achieve higher precision when compared with the-state-of-the-art method.