A bagging SVM to learn from positive and unlabeled examples

We consider the problem of learning a binary classifier from a training set of positive and unlabeled examples, both in the inductive and in the transductive setting. This problem, often referred to as PU learning, differs from the standard supervised classification problem by the lack of negative examples in the training set. It corresponds to an ubiquitous situation in many applications such as information retrieval or gene ranking, when we have identified a set of data of interest sharing a particular property, and we wish to automatically retrieve additional data sharing the same property among a large and easily available pool of unlabeled data. We propose a new method for PU learning with a conceptually simple implementation based on bootstrap aggregating (bagging) techniques: the algorithm iteratively trains many binary classifiers to discriminate the known positive examples from random subsamples of the unlabeled set, and averages their predictions. We show theoretically and experimentally that the method can match and even outperform the performance of state-of-the-art methods for PU learning, particularly when the number of positive examples is limited and the fraction of negatives among the unlabeled examples is small. The proposed method can also run considerably faster than state-of-the-art methods, particularly when the set of unlabeled examples is large. (C) 2013 Elsevier B.V. All rights reserved.