Can semi-supervised learning use all the data effectively? A lower bound perspective.
Alexandru TifreaGizem YüceAmartya SanyalFanny YangPublished in: CoRR (2023)
Keyphrases
- kernel methods
- semi supervised learning
- machine learning
- labeled data
- unlabeled data
- training data
- feature space
- semi supervised
- lower bound
- data analysis
- data sets
- co training
- learning models
- original data
- semi supervised classification
- training samples
- knowledge discovery
- data points
- active learning
- labeled training data
- label propagation
- domain adaptation
- labeled instances
- bayesian networks
- semi supervised learning methods
- information retrieval
- data distribution
- missing data
- prior knowledge
- pattern recognition
- reinforcement learning