Reducing Label Effort: Self-Supervised meets Active Learning.
Javad Zolfaghari BengarJoost van de WeijerBartlomiej TwardowskiBogdan C. RaducanuPublished in: CoRR (2021)
Keyphrases
- active learning
- labeling effort
- label noise
- active learning framework
- annotation effort
- unlabeled instances
- supervised learning
- random sampling
- semi supervised
- active learner
- labeled data
- selective sampling
- learning process
- multi label
- generalization error
- significantly reduced
- query by committee
- learning strategies
- learning algorithm
- batch mode active learning
- machine learning
- active learning strategies
- experimental design
- training examples
- unlabeled data
- training set
- genetic algorithm
- information retrieval