A new classifier based on information theoretic learning with unlabeled data.
Kyu-Hwa JeongJian-Wu XuDeniz ErdogmusJosé Carlos PríncipePublished in: Neural Networks (2005)
Keyphrases
- information theoretic
- unlabeled data
- supervised learning
- semi supervised learning
- learning algorithm
- labeled and unlabeled data
- active learning
- co training
- semi supervised
- information bottleneck
- mutual information
- labeled data
- training set
- learning process
- supervised learning algorithms
- prior knowledge
- learning problems
- multi view learning
- positive examples
- text classification
- machine learning
- class labels
- sample selection bias
- multi class
- decision trees
- unsupervised learning
- support vector machine
- labeled examples
- pairwise
- semi supervised classification
- reinforcement learning
- transductive learning
- training data
- image segmentation