Active learning without unlabeled samples: generating questions and labels using Monte Carlo Tree Search.
Sathish K. SankarPandiSpyros SamothrakisLuca CitiPeter BradyPublished in: IEEE BigData (2019)
Keyphrases
- unlabeled samples
- semi supervised
- monte carlo tree search
- active learning
- semi supervised learning
- labeled data
- labeled samples
- unlabeled data
- monte carlo
- evaluation function
- labeled and unlabeled data
- supervised learning
- semi supervised classification
- incremental learning
- pairwise
- training examples
- training samples
- labeled training data
- label propagation
- data distribution
- metric learning
- label information
- co training
- unsupervised learning
- game tree
- reinforcement learning methods
- data mining
- markov chain
- training set
- training data
- data sets
- temporal difference
- pairwise constraints
- class labels
- mixture model
- em algorithm
- high confidence
- learning algorithm
- machine learning