Bayesian Active Learning for Maximal Information Gain on Model Parameters.
Kasra ArnavazAasa FeragenOswin KrauseMarco LoogPublished in: ICPR (2020)
Keyphrases
- information gain
- active learning
- decision trees
- text categorization
- feature selection
- mutual information
- chi square
- chi squared
- semi supervised
- bayesian networks
- naive bayes
- decision tree learners
- correlation coefficient
- learning algorithm
- supervised learning
- occurrence frequency
- semi supervised learning
- learning process
- labeled data
- document frequency
- gain ratio
- unlabeled data
- generalization error
- machine learning
- relevance feedback
- cost sensitive
- logistic regression
- training set
- gini index