Looking into How Machine Learning is Used for Regression Testing Within the Agile Software Development Context

Abstract

Regression testing is essential for maintaining software reliability, especially in agile development environments marked by frequent code changes and iterative delivery cycles. Traditional regression testing methods often fall short in keeping up with the speed and complexity of modern development, leading to a growing need for smarter and more efficient solutions. In this context, Machine Learning (ML) has been developed as a transformational method, offering intelligent capabilities such as test case selection, fault prediction, and change impact analysis. These developments ensure that new code modifications do not adversely affect current functionality by automating and optimizing testing. This paper presents a comprehensive review of how ML is being applied to enhance regression testing in agile settings, highlighting current methodologies, categorizing state-of-the-art applications, and outlining the benefits in terms of accuracy, efficiency, and automation. The review aims to support researchers and practitioners in leveraging ML to develop more effective and Agile software development techniques for adaptive regression testing.