Gradient-based Novelty Detection Boosted by Self-supervised Binary Classification.
Jingbo SunLi YangJiaxin ZhangFrank LiuMahantesh HalappanavarDeliang FanYu CaoPublished in: CoRR (2021)
Keyphrases
- novelty detection
- binary classification
- multi class
- support vector
- learning problems
- cost sensitive
- multi label
- generalization error
- anomaly detection
- concept drift
- prediction accuracy
- class imbalance
- support vector machine
- binary classifiers
- ensemble methods
- data sets
- learning process
- data streams
- semi supervised
- information extraction
- upper bound
- prior knowledge
- training data
- image processing
- feature selection
- computer vision
- learning algorithm