HSDD: A hybrid sampling strategy for class imbalance in defect prediction data sets.
Muhammed Maruf ÖztürkAhmet ZenginPublished in: ICDIM (2016)
Keyphrases
- sampling strategy
- class imbalance
- sampling methods
- data sets
- defect prediction
- class distribution
- active learning
- concept drift
- cost sensitive
- random sampling
- training data
- minority class
- feature selection
- training set
- software repositories
- high dimensionality
- original data
- data streams
- software projects
- multi class
- high dimensional data
- naive bayes
- dimensionality reduction
- supervised learning
- pairwise
- machine learning