Active Model Selection for Positive Unlabeled Time Series Classification.
Shen LiangYanchun ZhangJiangang MaPublished in: ICDE (2020)
Keyphrases
- model selection
- unsupervised learning
- cross validation
- bayesian learning
- hyperparameters
- mixture model
- parameter estimation
- statistical learning
- model selection criteria
- variable selection
- statistical inference
- regression model
- machine learning
- error estimation
- sample size
- meta learning
- motion segmentation
- active learning
- training data
- generalization error
- selection criterion
- information criterion
- feature selection
- gaussian process
- unlabeled data
- parameter determination
- semi supervised learning
- supervised learning
- upper bound
- support vector machine
- prior knowledge