Evidential uncertainties on rich labels for active learning.
Arthur HoarauVincent LemaireArnaud MartinJean-Christophe DuboisYolande Le GallPublished in: CoRR (2023)
Keyphrases
- active learning
- annotation effort
- training examples
- unlabeled examples
- label noise
- labeling effort
- training set
- unlabeled instances
- semi supervised
- active learning framework
- ground truth labels
- active learner
- pool based active learning
- learning algorithm
- label propagation
- learning strategies
- selective sampling
- pairwise
- supervised learning
- learning process
- labeled examples
- labeled data
- labeled instances
- training data
- machine learning
- random sampling
- experimental design
- generalization error
- transfer learning
- semi supervised learning
- high level
- image labeling
- labeling scheme
- data mining
- multi label
- relevance feedback
- xml documents
- music emotion classification