A Contextual Bandit Approach for Stream-Based Active Learning.
Linqi SongJie XuPublished in: CoRR (2017)
Keyphrases
- active learning
- contextual bandit
- upper confidence bound
- data streams
- news recommendation
- sliding window
- training set
- learning algorithm
- selective sampling
- experimental design
- semi supervised
- machine learning
- learning strategies
- random sampling
- supervised learning
- relevance feedback
- semi supervised learning
- learning process
- image classification
- labeled data
- feature selection