Rescaling and other forms of unsupervised preprocessing introduce bias into cross-validation.
Amit MoscovichSaharon RossetPublished in: CoRR (2019)
Keyphrases
- cross validation
- preprocessing
- model selection
- support vector
- hyperparameters
- cross validated
- training set
- variable selection
- regularization parameter
- error estimates
- regression problems
- generalization error
- supervised learning
- semi supervised
- unseen data
- information criterion
- machine learning
- nearest neighbor
- nearest neighbor classifiers
- grid search
- data sets