Adversarial Active Learning in the Presence of Weak and Malicious Oracles.
Yan ZhouMurat KantarciogluBowei XiPublished in: PAKDD (Workshops) (2019)
Keyphrases
- active learning
- selective sampling
- random sampling
- learning algorithm
- learning strategies
- annotation effort
- training set
- semi supervised
- supervised learning
- unlabeled data
- test cases
- labeled data
- learning process
- semi supervised learning
- training examples
- relevance feedback
- multi agent
- machine learning
- label noise
- transfer learning
- experimental design
- malicious behavior
- decision trees
- pool based active learning