Using the one-versus-rest strategy with samples balancing to improve pairwise coupling classification.
Wieslaw ChmielnickiKatarzyna StaporPublished in: Int. J. Appl. Math. Comput. Sci. (2016)
Keyphrases
- training samples
- classification method
- pattern recognition
- training set
- classification systems
- classification algorithm
- machine learning
- feature extraction
- support vector
- classification accuracy
- support vector machine
- automatic classification
- avoid overfitting
- optimum path forest
- machine learning methods
- decision trees
- feature selection
- classification models
- supervised learning
- feature vectors
- classification process
- machine learning algorithms
- model selection
- text classification
- classification rules
- improve the classification accuracy