Leave-One-Out Cross-Validation Based Model Selection Criteria for Weighted LS-SVMs.
Gavin C. CawleyPublished in: IJCNN (2006)
Keyphrases
- leave one out cross validation
- model selection
- k nearest neighbour
- cross validation
- variable selection
- ls svm
- support vector machine
- selection criterion
- generalization error
- knn
- feature selection
- hyperparameters
- parameter estimation
- active learning
- information gain
- nearest neighbour
- regression model
- machine learning
- mutual information
- kernel learning
- feature space