MixPUL: Consistency-based Augmentation for Positive and Unlabeled Learning
This addresses the challenge of limited labeled data in PU learning, though it is incremental as it builds on existing methods with a new augmentation approach.
The paper tackles the problem of learning from positive and unlabeled data (PU learning) by proposing a data augmentation method based on consistency regularization, which achieves an averaged improvement in classification error from 16.49 to 13.09 on the CIFAR-10 dataset.
Learning from positive and unlabeled data (PU learning) is prevalent in practical applications where only a couple of examples are positively labeled. Previous PU learning studies typically rely on existing samples such that the data distribution is not extensively explored. In this work, we propose a simple yet effective data augmentation method, coined~\algo, based on \emph{consistency regularization} which provides a new perspective of using PU data. In particular, the proposed~\algo~incorporates supervised and unsupervised consistency training to generate augmented data. To facilitate supervised consistency, reliable negative examples are mined from unlabeled data due to the absence of negative samples. Unsupervised consistency is further encouraged between unlabeled datapoints. In addition,~\algo~reduces margin loss between positive and unlabeled pairs, which explicitly optimizes AUC and yields faster convergence. Finally, we conduct a series of studies to demonstrate the effectiveness of consistency regularization. We examined three kinds of reliable negative mining methods. We show that~\algo~achieves an averaged improvement of classification error from 16.49 to 13.09 on the CIFAR-10 dataset across different positive data amount.