Deep Active Learning for Sequence Labeling Based on Diversity and Uncertainty in Gradient.
Yekyung KimPublished in: CoRR (2020)
Keyphrases
- sequence labeling
- active learning
- conditional random fields
- dependency parsing
- structured prediction
- named entity recognition
- semi supervised
- supervised learning
- learning algorithm
- crf model
- markov models
- random sampling
- unlabeled data
- hedge detection
- information extraction
- hidden markov models
- learning process
- labeled and unlabeled data
- training examples
- labeled data
- transfer learning
- learning tasks
- conditional probabilities
- probabilistic inference
- higher order
- training set
- support vector
- information retrieval