نوع مقاله : مقاله پژوهشی
نویسندگان
1 کارشناسی ارشد، دانشگاه جامع امام حسین، تهران ایران
2 استادیار، دانشگاه جامع امام حسین (ع)، تهران، ایران
3 کارشناسی ارشد ، دانشگاه جامع امام حسین(ع) ، ایران، تهران
چکیده
کلیدواژه‌ها
موضوعات
عنوان مقاله [English]
نویسندگان [English]
A software defect, error, or failure is an incorrect or unexpected behavior in a software product. It results in undesired outcomes and behaves in unintended ways. This is a flaw in a software product that leads to unexpected performance. Predicting software defects involves the process of tracking defective components in the software and before starting the testing phases. In fact, the occurrence of defects is inevitable, but we must minimize the number of defects. This leads to a reduction in development time and cost, reduction in rework efforts, increased customer satisfaction, and software reliability. In this research, a multi-stage process has been used to predict software errors, where the pre-processing operation was performed to improve the data in the first stage. After the improvement operation, feature extraction operations were used to extract influential features. In this research, the wild horse algorithm was used to extract features. Finally, the decision tree algorithm was used to predict software errors. The evaluation metrics in this research are classification metrics, including accuracy, precision, recall, and F-measure. In this research, the standard dataset described in the base article was used for predicting software errors. Finally, a comparison was made between the proposed method and the base article method. The results showed that the performance of the wild horse algorithm (in the proposed method) is superior to the whale algorithm (in the base article) for feature selection because the wild horse algorithm was able to traverse the problem space appropriately. Additionally, the proposed algorithm utilized the theory of chaos to create diversity in the solutions. Therefore, the performance of the wild horse algorithm is superior to the whale algorithm in terms of comparison.
کلیدواژه‌ها [English]
)