Keywords: software quality assurance, software verification, artificial neural network, regression testing
Neural network based solution for regression testing optimization
UDC 004.054
DOI: 10.26102/2310-6018/2020.28.1.032
Regression testing is important task of retesting software systems after changes in the code of product to ensure that changes do not influence previously implemented functionality. Regression testing is run after a new version of software has been developed. Usually only limited subset of test cases is executed for a new version of software through restricted resources. This shows the problem of selection the most important regression test cases. To cope with limited resources, different regression testing techniques was developed to reduce the number of test cases to be executed. One of these techniques is test case prioritization based on neural network model. Such mechanism can collect data about code changes from Version Control System and use it as inputs for neural network. The outputs for such neural network model are regression tests' execution results. Groups of regression tests can be united by functionality under the test. Neural network model can be trained on real results during the phase of software developing. Trained neural network can detect the most important test cases for execution after each change in product code. Such technique can be used to guide the focus of the testing efforts.
Keywords: software quality assurance, software verification, artificial neural network, regression testing
For citation: Danilov A.D., Mugatina V.M. Neural network based solution for regression testing optimization. Modeling, Optimization and Information Technology. 2020;8(1). URL: https://moit.vivt.ru/wp-content/uploads/2020/02/DanilovMugatina_1_20_1.pdf DOI: 10.26102/2310-6018/2020.28.1.032 (In Russ).
Published 31.03.2020