A DCT based approach for detecting novelty and concept drift in data streams.
Morteza Zi HayatMahmoud Reza HashemiPublished in: SoCPaR (2010)
Keyphrases
- concept drift
- data streams
- novelty detection
- streaming data
- drifting concepts
- mining data streams
- sliding window
- data stream mining
- change detection
- evolving data streams
- data stream classification
- stream data
- drift detection
- incremental clustering
- data distribution
- non stationary
- classification algorithm
- batch learning
- stream mining
- ensemble classifier
- high speed data streams
- concept descriptions
- noisy data streams
- outlier detection
- machine learning
- small number
- support vector machine
- computer vision
- data sets