A Novel Approach to Predict the Defect Density in Software Application Using Linear Regression Algorithm

Aim: The aim of this project is to employ machine learning techniques to create a prediction model for fault density in software development, with a particular emphasis on the use of the Linear Regression algorithm. Materials and Methods: Two groups have participated in the current effort. In order to analyse the defect density of software applications, Group 1 refers to the case-based reasoning approach and Group 2 refers to the linear regression algorithm. Results: The defect density prediction of existing case-based reasoning methods achieves a 79 percent accuracy where the current method will increase its accuracy rate. Conclusion: This study highlights the effectiveness of the Linear Regression algorithm in defect density prediction, showcasing its potential to outperform traditional case-based methods.