Classification of positive and negative test cases using ensemble machine learning methods

In today’s IT industry, fast software delivery is becoming a standard and the race to the market adds up to the pressure. To ensure the quality of the developed features, quality assurance engineers apply both manual and automated techniques, and adapt their approach to optimize the process of verification. Machine learning can help significantly with the automation of some traditionally manual processes. In this research, we are showing how the classification of test cases into positive and negative can be done using ensemble machine learning methods, and whether those have advantages and better results over the basic machine learning methods. The best result was obtained by the Gradient Boosting Classifier with accuracy of 94%. However, since there is just a 1% difference in accuracy when compared to regular Decision Tree, it would not be necessary to use ensemble methods for this specific problem.