Classification algorithm sensitivity to training data with non representative attribute noise.
Michael V. ManninoYanjuan YangYoung RyuPublished in: Decis. Support Syst. (2009)
Keyphrases
- classification algorithm
- training data
- attribute noise
- training set
- class noise
- learning algorithm
- class labels
- naive bayes
- support vector machine
- noisy data
- agnostic learning
- concept drift
- classification rules
- classification accuracy
- supervised learning
- decision trees
- test data
- training samples
- training examples
- k nearest neighbor
- noise model
- data sets
- base learners
- unlabeled data
- knn
- test set
- data quality
- classification models
- prior knowledge
- probability estimation
- generalization error
- support vector
- feature selection
- classification noise
- bayesian networks
- classifier ensemble
- learning process
- training dataset
- multi class
- text classification
- semi supervised learning
- text categorization