Using Neural Networks to Schedule Automated Testing in an Agile Environment
Laurie L. Butgereit · 2019
Agile methodologies highly recommend the use of automated testing to ensure the quality of the software delivered to the customer. As a software project grows, however, the number of automated tests also grows. It is often the case that the automated tests start taking hours to execute. It would be beneficial that tests which have a high probability of failing be moved to the head of the test execution queue so that failures could be identified early in the test run. In this manner, programmers could begin remedial action on source classes where the test programs failed prior to the entire test run finishing executing. Prior research by the author used Weka and J48 algorithm in order to prioritize the execution of tests. The research in this paper compares the use of J48 and a multi-layer perceptron.