Plurality voting-based multiple classifier systems: statistically independent with respect to dependent classifier sets.
Mübeccel DemireklerHakan AltinçayPublished in: Pattern Recognit. (2002)
Keyphrases
- multiple classifier systems
- statistically independent
- majority voting
- dynamic classifier selection
- random variables
- classifier combination
- classification systems
- pattern recognition
- machine learning
- class conditional
- classifier ensemble
- decision trees
- ensemble methods
- classification accuracy
- support vector machine
- individual classifiers
- texture classification
- base classifiers
- binary classification
- training data
- face recognition
- computer vision
- neural network