Combining self-labeling and demand based active learning for non-stationary data streams.
Valerie VaquetFabian HinderJohannes BrinkrolfBarbara HammerPublished in: CoRR (2023)
Keyphrases
- non stationary
- active learning
- data streams
- concept drift
- active learning strategies
- adaptive algorithms
- streaming data
- sliding window
- empirical mode decomposition
- random sampling
- autoregressive
- semi supervised
- machine learning
- sample selection
- selective sampling
- white noise
- training set
- labeled data
- stock price
- stream data
- learning process
- learning algorithm
- temporal evolution
- batch mode
- data sets
- sensor networks
- multiresolution
- financial time series
- feature space
- image segmentation
- multi component
- labeling effort