A new concept drift detection method based on the ranking of features in a data stream.
Krzysztof WrobelPiotr PorwikTomasz OrczykBenjamin Mensah DadziePublished in: KES (2023)
Keyphrases
- concept drift
- detection method
- data streams
- feature detection
- streaming data
- drifting concepts
- data stream mining
- ensemble classifier
- detection algorithm
- non stationary
- evolving data streams
- stream data
- drift detection
- novelty detection
- data distribution
- mining data streams
- change detection
- face detection
- data stream classification
- feature extraction
- target concept
- classification algorithm
- sliding window
- benchmark datasets
- region detection
- feature space
- data sets
- text classification
- low dimensional
- feature set
- knn
- image analysis