Multi-label Learning by Hyperparameters Calibration for Treating Class Imbalance.
Andrés Felipe Giraldo-ForeroAndrés Felipe Cardona-EscobarAndrés Eduardo Castro-OspinaPublished in: HAIS (2018)
Keyphrases
- class imbalance
- hyperparameters
- random sampling
- active learning
- model selection
- cross validation
- support vector
- closed form
- bayesian inference
- bayesian framework
- class distribution
- maximum likelihood
- noise level
- em algorithm
- feature selection
- cost sensitive
- maximum a posteriori
- prior information
- multi label
- sample size
- concept drift
- machine learning
- unlabeled data
- training set
- training examples
- semi supervised
- regression problems
- missing values
- genetic algorithm ga
- text categorization
- generative model
- probabilistic model
- prior knowledge
- learning algorithm
- data sets