Contextual Bandit for Active Learning: Active Thompson Sampling.
Djallel BouneffoufRomain LarocheTanguy UrvoyRaphaël FéraudRobin AllesiardoPublished in: ICONIP (1) (2014)
Keyphrases
- active learning
- contextual bandit
- random sampling
- upper confidence bound
- sampling strategies
- active sampling
- selective sampling
- sample selection
- sampling methods
- stratified sampling
- relevance feedback
- unlabeled data
- semi supervised
- imbalanced class distribution
- sample size
- semi supervised learning
- sampling strategy
- learning algorithm
- natural language
- text categorization
- labeled data