Compensation of feature selection biases accompanied with improved predictive performance for binary classification by using a novel ensemble feature selection approach.
Ursula NeumannMona RiemenschneiderJan-Peter SowaTheodor BaarsJulia KälschAli CanbayDominik HeiderPublished in: BioData Min. (2016)
Keyphrases
- binary classification
- ensemble feature selection
- ensemble methods
- feature selection
- ensemble learning
- support vector
- multi class
- prediction accuracy
- feature subset
- decision trees
- benchmark datasets
- support vector machine
- cost sensitive
- learning problems
- genetic search
- generalization ability
- base classifiers
- machine learning methods
- class imbalance
- machine learning
- multi label
- generalization error
- text classification
- gene selection
- microarray
- text categorization
- unsupervised learning
- model selection
- feature selection algorithms
- bayesian networks
- learning algorithm
- neural network