Rethinking Class-Prior Estimation for Positive-Unlabeled Learning.
Yu YaoTongliang LiuBo HanMingming GongGang NiuMasashi SugiyamaDacheng TaoPublished in: ICLR (2022)
Keyphrases
- prior knowledge
- supervised learning
- positive examples
- learning systems
- learning process
- learning algorithm
- positive and unlabeled examples
- unsupervised learning
- online learning
- neural network
- reinforcement learning
- knowledge acquisition
- active learning
- support vector
- learning tasks
- inductive inference
- bayesian methods
- pattern languages
- statistical queries