جستجوی مقالات مرتبط با کلیدواژه « الگوریتم گرگ خاکستری » در نشریات گروه « پدافند غیرعامل »

A code smell is a superficial symptom that may indicate a deeper problem in the application. Code smell makes it difficult to maintain, develop and evolve the program. The presence of code smell definitely does not mean that the software is not working properly, but this issue may cause slow processing, increase the risk of failure and software errors. It is obvious that one of the effective ways to increase the quality of the software is to rebuild and rearrange the code, which has a direct relationship with the smell of the code. So far, a lot of research has been done in the field of identifying and removing code smells of software systems. However, among them, four types of code smell include; Long method, feature envy, god class and data class have attracted the most attention of researchers. Researchers use feature selection algorithms to increase the prediction accuracy of code smells and reduce data dimensions. In this article, the gray wolf algorithm has been used to determine the selected subset of optimal features. Selecting the feature makes the model simpler, improves the accuracy and also reduces the training time. Also, in order to identify and classify code smells, the recognition model based on three machine learning algorithms under the title; Nearest neighbor, decision tree and support vector machine are built. Finally, the final result of the model output is determined based on the majority voting mechanism. In this article, the improved version of the Fontana dataset has been used to evaluate the proposed method. Also, to evaluate the results of the proposed method, statistical criteria including precision, accuracy, recall and F criterion have been used. Finally, the results of the proposed method have been compared with the results of other related methods. The results obtained from the tests show that the proposed method has provided an acceptable performance compared to other methods.