Reconstructing Training Data From Real World Models Trained with Transfer Learning.
Yakir OzGilad YehudaiGal VardiItai AntebiMichal IraniNiv HaimPublished in: CoRR (2024)
Keyphrases
- transfer learning
- real world
- training data
- labeled data
- domain adaptation
- training set
- data sets
- learning algorithm
- multi task
- learning tasks
- target domain
- knowledge transfer
- semi supervised learning
- model selection
- supervised learning
- machine learning
- text classification
- machine learning algorithms
- active learning
- previously learned
- reinforcement learning
- training process
- labeled examples
- training and test data
- unlabeled data
- training examples
- prior knowledge
- cross domain
- test data
- probabilistic model
- manifold alignment