A new metric splitting criterion for decision trees.
Dan A. SimoviciSzymon JaroszewiczPublished in: Int. J. Parallel Emergent Distributed Syst. (2006)
Keyphrases
- decision trees
- predictive accuracy
- naive bayes
- training data
- decision tree induction
- selection criterion
- normalized mutual information
- random forest
- metric space
- distance measure
- metric learning
- decision tree learning
- rule induction
- similarity metric
- minimum description length
- feature construction
- distance metric
- feature space
- euclidean metric
- optimization criterion
- decision tree algorithm
- gini index
- machine learning
- constructive induction
- attribute selection
- information content
- rule sets
- classification rules
- logistic regression
- genetic algorithm
- information retrieval