Critical Gap Between Generalization Error and Empirical Error in Active Learning.
Yusuke KanebakoPublished in: WACV (2024)
Keyphrases
- generalization error
- active learning
- rademacher complexity
- uniform convergence
- training error
- expected error
- learning algorithm
- classification error
- training set
- linear classifiers
- target function
- supervised learning
- sample complexity
- binary classification
- machine learning
- random sampling
- query by committee
- error estimation
- semi supervised
- learning machines
- labeled data
- learning process
- risk minimization
- upper bound
- unlabeled data
- cost sensitive
- training examples
- generalization error bounds
- statistical learning theory
- semi supervised learning
- boosting algorithms
- conditional expectation
- data sets
- support vector machine
- test set
- experimental design
- theoretical analysis
- convergence rate