ProtoDA: Efficient Transfer Learning for Few-Shot Intent Classification.
Manoj KumarVarun KumarHadrien GlaudeCyprien de LichyAman AlokRahul GuptaPublished in: CoRR (2021)
Keyphrases
- pairwise
- transfer learning
- semi supervised
- labeled data
- text classification
- machine learning
- learning tasks
- similarity measure
- cross domain learning
- class labels
- semi supervised learning
- structure learning
- machine learning algorithms
- decision trees
- cross domain text classification
- label information
- feature vectors
- transfer knowledge
- cross domain
- knowledge transfer
- supervised learning
- active learning
- neural network
- multi task learning
- support vector
- feature extraction
- databases
- support vector machine
- learning algorithm
- co training
- multi task
- feature selection
- reinforcement learning
- text categorization
- target domain
- information retrieval
- unlabeled data
- manifold alignment
- feature space
- text mining