Efficient Model Selection for Regularized Classification by Exploiting Unlabeled Data.
Georgios BalikasIoannis PartalasÉric GaussierRohit BabbarMassih-Reza AminiPublished in: IDA (2015)
Keyphrases
- model selection
- unlabeled data
- cross validation
- semi supervised learning
- labeled data
- semi supervised
- supervised learning
- text classification
- co training
- semi supervised classification
- labeled and unlabeled data
- unsupervised learning
- class labels
- machine learning
- training set
- hyperparameters
- active learning
- training data
- selection criterion
- feature selection
- mixture model
- text categorization
- sample size
- regression model
- gaussian process
- transfer learning
- generalization error
- real world
- least squares
- decision boundary
- classification accuracy
- training examples
- supervised classification
- domain adaptation
- variable selection
- learning algorithm
- labeled training data
- learning problems
- pairwise
- reinforcement learning
- multi label
- training samples
- generative model
- data points
- feature space
- data mining