Online query by committee for active learning from drifting data streams.
Bartosz KrawczykMichal WozniakPublished in: IJCNN (2017)
Keyphrases
- query by committee
- active learning
- data streams
- random sampling
- online learning
- supervised learning
- concept drift
- uncertainty sampling
- sliding window
- sample complexity
- learning algorithm
- machine learning
- data sets
- semi supervised
- generalization error
- data distribution
- training set
- learning strategies
- learning process
- cost sensitive
- experimental design
- labeled data
- training examples
- upper bound
- logistic regression
- theoretical analysis
- sampling algorithm
- feature selection
- unlabeled data