Novel Automatic Filter-Class Feature Selection for Machine Learning Regression.
Morten Gill WollsenJohn HallamBo Nørregaard JørgensenPublished in: INNS Conference on Big Data (2016)
Keyphrases
- feature selection
- machine learning
- model selection
- text classification
- nonlinear filters
- feature set
- supervised feature selection
- class separability
- reproducing kernel hilbert space
- regression problems
- machine learning methods
- decision trees
- support vector
- mutual information
- pattern recognition
- learning machines
- preprocessing step
- semi automatic
- unsupervised learning
- support vector machine
- kernel methods
- microarray data
- cross validation
- classification and regression problems
- information extraction
- learning algorithm
- dimensionality reduction
- data mining
- wrapper method
- feature selection and classification
- feature generation
- feature ranking
- multi class
- discriminative features
- selected features
- information gain
- learning tasks