Boosting Multitask Decomposition: Directness, Sequentiality, Subsampling, Cross-Gradients.
András MillinghofferMátyás AntalMárk MarosiAndrás FormanekAndrás AntosPéter AntalPublished in: AIME (1) (2024)
Keyphrases
- multi task
- multitask learning
- feature selection
- multi task learning
- ensemble methods
- binary classification problems
- learning tasks
- multiple tasks
- multi class
- base learners
- gaussian processes
- learning problems
- multiclass support vector machines
- transfer learning
- base classifiers
- maximum margin
- machine learning
- mutual information
- multiclass classification
- learning algorithm
- information gain
- text categorization
- binary classification
- weak classifiers
- pairwise
- support vector
- learning environment
- information retrieval