Handling over-fitting in test cost-sensitive decision tree learning by feature selection, smoothing and pruning.
Tao WangZhenxing QinZhi JinShichao ZhangPublished in: J. Syst. Softw. (2010)
Keyphrases
- cost sensitive
- decision tree learning
- feature selection
- multi class
- naive bayes
- decision trees
- misclassification costs
- class imbalance
- information gain
- meta learning
- cost sensitive classification
- support vector machine
- text categorization
- class distribution
- attribute values
- active learning
- mutual information
- classification accuracy
- test data
- ensemble methods
- constructive induction
- machine learning
- support vector machine svm
- feature set
- decision tree induction
- learning algorithm
- knn