Software Defect Sentry using Ensemble Learning

These days, software is ingrained in every part of our life, so it's highly likely to be flawed. Software Defect Prediction (SDP) predicts both front-end and back-end flaws, among other technical problems. To predict bugs, all we have to do is use various Ensemble Learning techniques. Software Defect Prediction (SDP) is one of the most helpful techniques in the Testing Phase of the Software Development Life Cycle (SDLC). It identifies the modules that require testing because of their error- prone nature. This allows for efficient use of the testing resources without going above permitted limits. In the rapidly evolving world of software development, making sure a software system is of high quality is essential. Program testing is one of the most important techniques for guaranteeing program quality. It was found that the cost of testing accounted for almost half of the project's overall expenses. Effective and efficient software testing makes use of the least amount of software resources. Designing a procedure that may effectively carry out testing while also utilizing the fewest resources feasible for the project is therefore essential. Software Defect Sentry was thus developed. The goal of our Sentry is to find as many software bugs as it can. The world is always moving toward making important judgments based on facts. Therefore, we will do machine learning analysis on datasets that are made publicly available in this project in order to achieve the best level of accuracy feasible. The primary objective of the study is to apply several machine learning algorithms to the datasets and identify which methods produce the best outcomes. To be more precise, we demonstrated an ensemble learning model and compared the performance of KNN, Decision tree, SVM, and Naive Bayes on multiple datasets with the Ensemble technique. We also measured F1-score, accuracy, precision, and recall.