Sample size vs. bias in defect prediction.
Foyzur RahmanDaryl PosnettIsrael HerraizPremkumar T. DevanbuPublished in: ESEC/SIGSOFT FSE (2013)
Keyphrases
- sample size
- defect prediction
- variance reduction
- software repositories
- software projects
- model selection
- random sampling
- upper bound
- covariance matrix
- small sample size
- confidence intervals
- small sample
- experimental design
- worst case
- pac learning
- statistical power
- vc dimension
- random sample
- number of training samples
- small samples
- progressive sampling
- lower bound
- high level
- software development
- data streams
- address these issues
- decision trees
- historical data
- raw data
- software systems
- source code