Leave-One-Out Cross-Validation Based Model Selection for Manifold Regularization.
Jin YuanYanming LiChengliang LiuXuan F. ZhaPublished in: ISNN (1) (2010)
Keyphrases
- leave one out cross validation
- model selection
- cross validation
- semi supervised learning
- hyperparameters
- sample size
- regression model
- variable selection
- generalization error
- machine learning
- semi supervised
- gaussian process
- feature selection
- selection criterion
- unsupervised learning
- principal component analysis
- data mining
- face recognition
- dimension reduction
- convex optimization
- labeled and unlabeled data