Backdoor Attacks Against Transfer Learning With Pre-Trained Deep Learning Models.
Shuo WangSurya NepalCarsten RudolphMarthie GroblerShangyu ChenTianle ChenPublished in: IEEE Trans. Serv. Comput. (2022)
Keyphrases
- learning models
- transfer learning
- pre trained
- learning tasks
- semi supervised learning
- training data
- labeled data
- learning algorithm
- machine learning algorithms
- machine learning
- training examples
- learning problems
- active learning
- reinforcement learning
- multi task
- unlabeled data
- semi supervised
- text categorization
- control signals
- classification models
- collaborative filtering
- loss function
- text classification
- conditional random fields
- supervised learning
- unsupervised learning
- decision trees
- data sets
- multi label
- image processing
- training samples
- hidden markov models
- prior knowledge