Hellinger distance decision trees are robust and skew-insensitive.
David A. CieslakT. Ryan HoensNitesh V. ChawlaW. Philip KegelmeyerPublished in: Data Min. Knowl. Discov. (2012)
Keyphrases
- decision trees
- decision tree induction
- genetic algorithm
- computationally efficient
- distance measure
- machine learning
- decision rules
- naive bayes
- computer vision
- data mining
- association rules
- bayesian networks
- distance function
- learning algorithm
- partial occlusion
- robust estimation
- rule induction
- mahalanobis distance
- kullback leibler