A Comparative Study of Four Metaheuristic Algorithms for Test Suite Reduction

Test Suite Reduction (TSR) is required in software testing to minimize test suite size while making sure full requirement coverage. The main challenge of TSR lies in achieving high reduction without sacrificing coverage or increasing computational time. Therefore, the goal of this study is identifying the better optimization algorithm for TSR, particularly when balancing reduction rate, coverage, and execution time. This study compares the performance of four algorithms which include Rat Swarm Optimization, Genetic Algorithm, Particle Swarm Optimization, and Grey Wolf Optimizer in optimizing TSR problem. The algorithms were evaluated using multiple benchmark datasets over 10 independent runs, with metrics including reduction percentage, requirement coverage, and execution time. The experimental findings show that RSO outperforms other algorithms, achieving the top percentage test suite reduction under keeping coverage at its peak and way faster execution times. GA performed similarly but demands a lot of computing power, while PSO and GWO have trouble finding the best solution quickly and handling larger problems. This data validates RSO as a promising technique for TSR, where future studies will be concentrated on integrating it into a multifactorial optimization framework to enhance efficiency in multi-task TSR scenario.