A variance maximization criterion for active learning.
Yazhou YangMarco LoogPublished in: Pattern Recognit. (2018)
Keyphrases
- active learning
- pool based active learning
- selection criterion
- random sampling
- machine learning
- training examples
- selective sampling
- relative entropy
- learning algorithm
- semi supervised
- supervised learning
- labeled data
- semi supervised learning
- minimum description length
- standard deviation
- cost sensitive
- sample selection
- posterior marginals
- music emotion classification
- generalization error
- training set
- data sets
- correlation coefficient
- unlabeled data
- variance reduction
- optimization criterion
- objective function
- information retrieval
- annotation effort
- neural network
- batch mode active learning
- imbalanced data classification
- database