Investigating Exploratory Capabilities of Uncertainty Sampling using SVMs in Active Learning.
Dominik LangDaniel KottkeGeorg KremplMyra SpiliopoulouPublished in: AL@iKNOW (2016)
Keyphrases
- uncertainty sampling
- active learning
- cost sensitive
- support vector
- experimental design
- training set
- binary classification
- multi class
- random sampling
- machine learning
- query by committee
- misclassification costs
- support vector machine
- naive bayes
- learning strategies
- probability estimates
- feature selection
- labeled data
- ensemble members
- logistic regression
- generalization error
- training examples
- support vector machine svm
- semi supervised
- svm classifier
- class imbalance
- supervised learning
- learning process
- learning algorithm
- maximum entropy
- class distribution
- semi supervised learning
- training data
- decision trees
- sample complexity