Nested cross-validation when selecting classifiers is overzealous for most practical applications.
Jacques WainerGavin C. CawleyPublished in: Expert Syst. Appl. (2021)
Keyphrases
- cross validation
- support vector
- training set
- nearest neighbor classifiers
- unseen data
- model selection
- binary classifiers
- roc analysis
- variable selection
- hyperparameters
- cross validated
- training data
- grid search
- regularization parameter
- decision trees
- generalization error
- test set
- naive bayes
- meta learning
- feature selection
- error estimates
- svm classifier
- training samples
- kernel function
- data sets
- support vector machine
- rbf kernel
- least squares
- logistic regression
- supervised classification
- test data
- ls svm
- sample size
- leave one out cross validation
- information criterion
- multiclass problems
- roc curve
- classification accuracy
- feature extraction
- selection algorithm