A Classification and Novel Class Detection Algorithm for Concept Drift Data Stream Based on the Cohesiveness and Separation Index of Mahalanobis Distance.
Xiangjun LiYong ZhouZiyan JinPeng YuShun ZhouPublished in: J. Electr. Comput. Eng. (2020)
Keyphrases
- labeled data
- concept drift
- class labels
- detection algorithm
- classification algorithm
- data streams
- mahalanobis distance
- data stream classification
- supervised learning
- streaming data
- data distribution
- outlier detection
- training examples
- learning algorithm
- sliding window
- data stream mining
- metric learning
- change detection
- evolving data streams
- moving objects
- data sets
- pattern recognition
- covariance matrix
- support vector machine
- euclidean distance
- non stationary
- noise model
- distance metric
- nearest neighbor
- support vector
- computer vision
- index structure