IJEDRCP1403031 International Jou A Methodology f Defect

Abstract — Predicting defect-prone software components is an economically important activity. Software defect prediction work focuses on three ways 1) Estimating the number of defects remaining in software systems, 2) Discovering defect associations, and 3) Classifying the defect proneness of software components. The software defect prediction that supports both unbiased and comprehensive comparison between competing prediction systems. This methodology is comprised of 1) scheme evaluation and 2) defect prediction components. The scheme evaluation analyzes the prediction performance of competing learning schemes for given historical data sets. The defect predictor builds models according to the evaluated learning scheme and predicts software defects with new different learning schemes are evaluated and best one is selected. In the prediction stage the best learning scheme is used to build a predictor with all historical data and the predictor is classifies the defect-proneness of software components into two classes, defec