Training Deep Nets with Imbalanced and Unlabeled Data.
Jeffrey BerryIan R. FaselLuciano FadigaDiana ArchangeliPublished in: INTERSPEECH (2012)
Keyphrases
- unlabeled data
- number of labeled examples
- labeled data
- semi supervised learning
- labeled training data
- training set
- supervised learning
- semi supervised
- labeled data for training
- training examples
- class distribution
- active learning
- labeled examples
- co training
- semi supervised classification
- training data
- text classification
- learning algorithm
- text categorization
- labeled and unlabeled data
- training and test data
- data points
- training samples
- positive examples
- unlabeled training data
- labeled instances
- supervised methods
- training process
- transfer learning
- domain adaptation
- small set of labeled
- test set
- data sets
- class imbalance
- class labels
- prior knowledge
- labeling effort
- knowledge discovery
- supervised learning algorithms
- support vector machine
- sample selection bias
- high dimensional
- pairwise
- similarity measure
- neural network