Attribute Relevance in Multiclass Data Sets Using the Naive Bayes Rule.
José Martínez SotocaJosé Salvador SánchezFiliberto PlaPublished in: ICPR (3) (2004)
Keyphrases
- multi class
- naive bayes
- data sets
- training data
- feature selection
- decision trees
- class probabilities
- base classifiers
- text classification
- classification accuracy
- cost sensitive
- naive bayes classifier
- logistic regression
- bayesian network classifiers
- support vector machine
- multiclass classification
- text categorization
- pairwise
- probability estimation
- binary classification
- classification algorithm
- bayesian networks
- multi class classification
- error correcting output codes
- training set
- information gain
- attribute values
- binary classifiers
- classification rules
- multi class problems
- rule sets
- misclassification costs
- probability estimates
- machine learning
- neural network
- class distribution
- training samples
- class imbalance
- high dimensional data
- supervised learning