MSSBoost: A new multiclass boosting to semi-supervised learning.
Jafar TanhaPublished in: Neurocomputing (2018)
Keyphrases
- semi supervised learning
- multi class
- boosting algorithms
- base classifiers
- multiclass classification
- cost sensitive
- semi supervised
- unlabeled data
- labeled data
- feature selection
- unsupervised learning
- machine learning
- supervised learning
- support vector machine
- kernel machines
- binary classification problems
- pairwise
- training data
- multi class classification
- binary classification
- text categorization
- learning problems
- learning models
- co training
- active learning
- learning algorithm
- weak classifiers
- ensemble methods
- transfer learning
- multi task
- binary classifiers
- reinforcement learning
- improve the classification accuracy
- decision boundary
- multiple kernel learning
- loss function
- error correcting output codes