Anticipating Software Fault Proneness using Classifier Ensemble: An Optimize Approach

Bushra Hamid, Eisa bin Abdullah Aleissa, and Abdul Rauf


Defect Prediction, Genetic Algorithm, Optimization, Majority Weighted Voting, Classifier Ensembles


Prediction of software faults earlier in software development life cycle makes testing process more effective. This enables software testers to concentrate on the modules that are being predicted as defective ones. By doing so, testing effort can be greatly reduced and renders the development of quality product. Different approaches have been proposed so far to predict software module as faulty or non-faulty (defective or non-defective). No doubt existing approaches perform well however there is still a great room for improvement. In this paper, we have proposed a novel approach for predicting software modules, based on combination of different machine learning algorithms and optimizing results by using Genetic algorithm. Performance measures used for evaluation purpose are accuracy of particular classifier in predicting faulty/non faulty module, probability of detection (Pd) and probability of false alarm (Pf). We have made a comparative analysis of our proposed approach with other existing defect prediction approaches.

Important Links:

Go Back