Cost-sensitive meta-learning classifiers: MEPAR-miner and DIFACONN-miner.
Sinem KullukLale ÖzbakirPinar TapkanAdil BaykasogluPublished in: Knowl. Based Syst. (2016)
Keyphrases
- meta learning
- cost sensitive
- base classifiers
- naive bayes
- inductive learning
- decision trees
- cost sensitive learning
- model selection
- learning tasks
- multi class
- machine learning algorithms
- classification algorithm
- misclassification costs
- cost sensitive classification
- class distribution
- feature selection
- active learning
- machine learning
- data mining
- class imbalance
- machine learning methods
- pairwise
- training data
- boosting algorithms
- unsupervised learning
- k nearest neighbor
- supervised learning
- support vector machine
- training set
- reinforcement learning