Software Defect Prediction Based Ensemble Approach

Software systems have grown significantly and in complexity. As a result of these qualities, preventing software faults is extremely difficult. defect prediction (SDP) can assist developers finding potential bugs reducing maintenance costs. When it comes to lowering costs assuring quality, SDP plays critical role development. result, automatically forecasting the number errors modules important, may allocating limited resources more efficiently. Several methods for detecting addressing such flaws at low cost been offered. These approaches, on other hand, need be improved terms performance. Therefore this paper, two deep learning (DL) models Multilayer preceptor (MLP) neural network (DNN) are proposed. The proposed approaches combine newly established Whale optimization algorithm (WOA) with complementary Firefly (FA) establish emphasized metaheuristic search EMWS algorithm, which selects fewer but closely related representative features. To find best-implemented classifier achievement measurement factor, classifiers were applied five PROMISE repository datasets. compared existing methods, technique outperforms, 0.91% JM1 dataset, 0.98% accuracy KC2 PC1 0.93% MC2 0.92% KC3.