Abstract:
"Techniques for selecting test cases shorten testing duration and cost, but they could omit some
crucial tests that can detect vulnerabilities. Test case prioritization considers all test cases and
performs them up until resources are depleted or all test cases have been executed, while
continuously focusing on the most important ones. Over time, machine learning has gained
popularity in software engineering and is now utilized to solve many different problems.
Learning concerns can include issues with software development and maintenance, and
machine learning techniques have been very effective in resolving these problems. Machine
learning approaches have been used to solve the test case prioritization problem, extending its
range of use.
This project will involve conducting research on test case prioritization using machine learning
techniques that have been considered in the past. Additionally, the project's primary focus will
be on the gaps and problems in earlier studies in the relevant field. This project proposal
emphasizes the problem that will be addressed, the novelty of the research idea, and the
research gap by using the pertinent evidence from previous research evaluations. At the
conclusion of this research, methodology and delivery plans are also covered.
A supervised learning approach with a few machine learning algorithms was used for test case
priority classification. Logistic Regression, SVM, Random Forest, Decision Tree, and KNN
were tested separately to evaluate the performance of each model. To enhance the accuracy of
the final classification model, an ensemble learning technique was used. From the ensemble
method, a Voting classifier was used with the above-mentioned ML algorithms to build the
final test case priority classifier."
