Classification of the drifting data streams using heterogeneous diversified dynamic class-weighted ensemble.
Martin SarnovskyMichal KolarikPublished in: PeerJ Comput. Sci. (2021)
Keyphrases
- data streams
- mining concept drifting data streams
- concept drift
- ensemble classifier
- concept drifting data streams
- final classification
- training set
- classification accuracy
- support vector machine
- class labels
- multi class classification
- streaming data
- multiclass classification
- pattern recognition
- support vector machine svm
- data stream classification
- sliding window
- classification algorithm
- classifier ensemble
- support vector
- training data
- nearest neighbour
- binary classification problems
- majority voting
- k nearest neighbour
- class specific
- drifting concepts
- neural network
- feature selection
- text classification
- data sets
- regression problems
- anytime classification
- feature extraction
- classification method
- machine learning algorithms
- model selection
- weak learners
- multiple classifiers
- imbalanced data
- sensor networks
- feature set
- multiple classifier systems
- individual classifiers
- multiple classes
- intra class
- data distribution
- ensemble methods
- generalization ability
- ensemble learning