Using unsupervised incremental learning to cope with gradual concept drift.
David HadasGalit YovelNathan IntratorPublished in: Connect. Sci. (2011)
Keyphrases
- incremental learning
- concept drift
- batch learning
- data stream classification
- semi supervised
- supervised learning
- novelty detection
- data streams
- non stationary
- streaming data
- classification algorithm
- drifting concepts
- incremental learning algorithm
- unsupervised learning
- drift detection
- incremental clustering
- data distribution
- fuzzy artmap
- learning process
- semi supervised learning
- evolving data streams
- training set
- change detection
- batch mode
- data stream mining
- learning environment
- training data
- learning tasks
- computer vision
- class labels
- labeled data
- feature extraction
- decision trees