Dist-PU: Positive-Unlabeled Learning from a Label Distribution Perspective

TL;DR

This paper introduces a novel positive-unlabeled learning method that leverages label distribution consistency and regularization techniques to improve classification performance without negative labels.

Contribution

It proposes a label distribution perspective for PU learning, aligning predicted and true label distributions, and incorporates entropy minimization and Mixup regularization.

Findings

Outperforms existing methods on benchmark datasets.

Effectively mitigates confirmation bias in PU learning.

Demonstrates robustness across different class priors.

Abstract

Positive-Unlabeled (PU) learning tries to learn binary classifiers from a few labeled positive examples with many unlabeled ones. Compared with ordinary semi-supervised learning, this task is much more challenging due to the absence of any known negative labels. While existing cost-sensitive-based methods have achieved state-of-the-art performances, they explicitly minimize the risk of classifying unlabeled data as negative samples, which might result in a negative-prediction preference of the classifier. To alleviate this issue, we resort to a label distribution perspective for PU learning in this paper. Noticing that the label distribution of unlabeled data is fixed when the class prior is known, it can be naturally used as learning supervision for the model. Motivated by this, we propose to pursue the label distribution consistency between predicted and ground-truth label distributions, which is formulated by aligning their expectations. Moreover, we further adopt the entropy minimization and Mixup regularization to avoid the trivial solution of the label distribution consistency on unlabeled data and mitigate the consequent confirmation bias. Experiments on three benchmark datasets validate the effectiveness of the proposed method.Code available at: https://github.com/Ray-rui/Dist-PU-Positive-Unlabeled-Learning-from-a-Label-Distribution-Perspective.