Active poisoning: efficient backdoor attacks on transfer learning-based brain-computer interfaces.
Xue JiangLubin MengSiyang LiDongrui WuPublished in: Sci. China Inf. Sci. (2023)
Keyphrases
- learning algorithm
- transfer learning
- learning tasks
- machine learning algorithms
- brain computer interface
- labeled data
- machine learning
- active learning
- knowledge transfer
- cross domain
- text classification
- unlabeled data
- multi task learning
- multi task
- target domain
- data mining
- transfer knowledge
- manifold alignment
- semi supervised learning
- text categorization
- information extraction
- object recognition