Extreme Value Theory for Open Set Classification - GPD and GEV Classifiers.
Edoardo VignottoSebastian EngelkePublished in: CoRR (2018)
Keyphrases
- extreme value theory
- support vector
- classification systems
- classification algorithm
- classification method
- training set
- classification process
- classification models
- class labels
- decision trees
- supervised classification
- svm classifier
- classifier combination
- feature selection
- classification procedure
- nearest neighbor classifier
- machine learning algorithms
- pattern recognition
- final classification
- optimum path forest
- classification rate
- accurate classification
- multi category
- individual classifiers
- classification decisions
- improves the classification accuracy
- probabilistic classifiers
- rule based classifier
- feature set
- classification accuracy
- naive bayes classifier
- training samples
- roc analysis
- multiple classifiers
- combining classifiers
- training data
- imbalanced data sets
- majority voting
- k nearest neighbour
- decision boundary
- support vector machine
- machine learning methods
- feature space
- binary classifiers
- classification schemes
- accurate classifiers
- text classification
- correctly classified
- support vector machine svm
- data stream classification
- learning algorithm
- discriminant functions
- discriminative classifiers
- classification trees
- fold cross validation
- multi layer perceptron
- supervised learning
- co training
- extracted features
- naive bayes