Homogeneous Transfer Active Learning for Time Series Classification.
Patrick Kinyua GikundaNicolas JouandeauPublished in: ICMLA (2021)
Keyphrases
- relevance feedback
- active learning
- transfer learning
- learning algorithm
- knowledge transfer
- semi supervised
- machine learning
- random sampling
- learning strategies
- labeled data
- unlabeled data
- selective sampling
- semi supervised learning
- online learning
- supervised learning
- nearest neighbor
- sampling methods
- accurate classifiers
- training set
- data sets
- small number
- dynamic time warping
- learning process
- experimental design
- pool based active learning
- generalization error
- class imbalance
- annotation effort