Identifying mislabeled training data with the aid of unlabeled data.
Donghai GuanWeiwei YuanYoung-Koo LeeSungyoung LeePublished in: Appl. Intell. (2011)
Keyphrases
- unlabeled data
- training data
- labeled data
- semi supervised learning
- semi supervised
- labeled training data
- supervised learning
- co training
- labeled data for training
- training set
- labeled examples
- active learning
- training examples
- text categorization
- semi supervised classification
- text classification
- class labels
- learning algorithm
- labeled and unlabeled data
- positive examples
- number of labeled examples
- classification accuracy
- class distribution
- data sets
- test data
- domain adaptation
- labeled instances
- label propagation
- label information
- prior knowledge
- data points
- transfer learning
- supervised learning algorithms
- test set
- small set of labeled
- negative examples
- training process
- training and test data
- training samples
- training dataset
- knn
- learning process
- target domain
- feature extraction
- decision trees