Combining Classifiers Using Dependency-Based Product Approximation with Bayes Error Rate.
Hee-Joong KangPublished in: Multiple Classifier Systems (2004)
Keyphrases
- bayes error rate
- combining classifiers
- conditional entropy
- error rate
- lower bound
- classifier combination
- bayes error
- information theory
- multiple classifiers
- combining multiple
- naive bayes
- majority voting
- base classifiers
- combination of multiple classifiers
- objective function
- mutual information
- decision trees
- test set
- data sets
- multi class
- knn
- training set
- pattern recognition
- computer vision