Comparing the performance of quantum-inspired evolutionary algorithms for the solution of software requirements selection problem

Abstract

ContextIn requirements engineering phase of the software development life cycle, one of the main concerns of software engineers is to select a set of software requirements for implementation in the next release of the software from many requirements proposed by the customers, while balancing budget and customer satisfaction. ObjectiveTo analyse the efficacy of Quantum-inspired Elitist Multi-objective Evolutionary Algorithm (QEMEA), Quantum-inspired Multi-objective Differential Evolution Algorithm (QMDEA) and Multi-objective Quantum-inspired Hybrid Differential Evolution (MQHDE) in solving the software requirements selection problem. MethodThe paper reports on empirical evaluation of the performance of three quantum-inspired multi-objective evolutionary algorithms along with Non-dominated Sorting Genetic Algorithm-II (NSGA-II). The comparison includes the obtained Pareto fronts, the three performance metrics – Generational Distance, Spread and Hypervolume, attained boundary solutions, and size of the Pareto front. ResultsThe results reveal that MQHDE outperformed other methods in producing high quality solutions; while QMDEA is able to produce well distributed solutions with extreme boundary solutions. ConclusionThe hybridization of Differential Evolution with Genetic Algorithms coupled with quantum computing concepts (MQHDE) provided a means to effectively balance the two issues of multi-objective optimization - convergence and diversity.