Refining the Wrapper Approach - Smoothed Error Estimates for Feature Selection.
Loo-Nin TeowHaifeng LiuHwee Tou NgEric YapPublished in: ICML (2002)
Keyphrases
- feature selection
- error estimates
- error estimation
- cross validation
- model selection
- mutual information
- feature ranking
- support vector
- text categorization
- error analysis
- feature space
- support vector machine
- feature set
- wrapper method
- dimensionality reduction
- multi class
- knn
- feature selection algorithms
- provably correct
- classification accuracy
- multi task
- method for feature selection
- discriminative features
- machine learning
- text classification
- generalization error
- information gain
- feature subset
- microarray data
- learning algorithm
- selected features
- irrelevant features
- high dimensionality
- data mining