PULNS: Positive-Unlabeled Learning with Effective Negative Sample Selector

Positive-unlabeled learning (PU learning) is an important case of binary classification where the training data only contains positive and unlabeled samples. The current state-of-the-art approach for PU cost-sensitive approach, which casts as a problem relies on unbiased risk estimator correcting bias introduced by However, this requires knowledge class prior subject to potential label noise. In paper, we propose novel dubbed PULNS, equipped with effective negative sample selector, optimized reinforcement learning. Our PULNS employs selector agent responsible selecting samples from data. While selected, likely can be used improve classifier, performance classifier also reward through REINFORCE algorithm. By alternating updates both improved. Extensive experimental studies 7 real-world application benchmarks demonstrate that consistently outperforms methods in learning, our results confirm effectiveness underlying PULNS.