Prediction of software defects using deep learning with improved cuckoo search algorithm

Abstract

Summary The software project model needs a defect prediction model to find defect‐prone file software systems. The fault‐prone model prediction, predicting bugs, and bug removal can undertake the software industry to achieve software quality. Therefore, automatically forecasting the number of errors in software modules is important, and it may assist developers in allocating limited resources more efficiently. Several methods for detecting and repairing such flaws at a low cost have been offered. These approaches, on the other hand, need to be significantly improved in terms of performance. Hence in this article, we implement an ensemble technique for the software defect prediction and prediction of the software bug. Also, we proposed a hybrid technique to predict several defects in the software system. The proposed approach uses principle component analysis for feature extraction which is to improve further performance and control the optimization problem. Classifiers were applied to five PROMISE datasets to determine the greatest implemented classifier with respect to the prediction achievement measuring factor. Our proposed model yields greater results on solving defect prediction problems and showing enhancement toward the existing model.