Improving Classifier Performance by Using Fictitious Training Data? A Case Study.
Ralf SteckingKlaus B. SchebeschPublished in: OR (2007)
Keyphrases
- training data
- class distribution
- learning algorithm
- data sets
- decision trees
- case study
- training set
- prior knowledge
- machine learning
- classification accuracy
- supervised learning
- test data
- training process
- unlabeled data
- training samples
- input data
- test set
- test bed
- cost sensitive
- roc curve
- training instances
- class imbalance
- error rate
- training examples
- labeled data
- object detection
- expert systems