Learning a robust foundation model against clean-label data poisoning attacks at downstream tasks.
Ting ZhouHanshu YanBo HanLei LiuJingfeng ZhangPublished in: Neural Networks (2024)
Keyphrases
- hidden variables
- probabilistic model
- network structure
- data sets
- active learning
- accurate models
- data sources
- supervised learning
- data collection
- learned models
- experimental data
- test data
- partially labeled
- multiple tasks
- multi task learning
- multi task
- learning models
- learning tasks
- statistical methods
- missing data
- computational model
- training data
- social networks