Improving Software Testing Process Using Fault Prediction Based on the Combination of Genetic Algorithm and Particle Swarm Optimization

Software testing is a critical process aimed at identifying defects, ensuring compliance with existing requirements, and verifying compatibility with customer hardware (environment). To reduce costs and enhance accuracy, predicting software defects before the testing phase becomes essential. Through predictive approaches, approximately ۹۵% of defects can be anticipated before reaching the testing stage. In this study, a hybrid approach combining Particle Swarm Optimization (PSO) and Genetic Algorithm (GA) is employed for feature selection in software defect prediction. This is followed by a decision tree classifier designed to achieve comprehensive defect coverage. The results demonstrate that the proposed method achieves an accuracy exceeding ۹۰% and outperforms baseline methods and previous studies in terms of quality.

Keywords:

Software Testing , Fault Prediction , Feature Selection , Genetic Algorithm , Particle Swarm Optimization