期刊名称：Journal of Theoretical and Applied Information Technology
印刷版ISSN：1992-8645
电子版ISSN：1817-3195
出版年度：2021
卷号：99
期号：3
页码：569
出版社：Journal of Theoretical and Applied
摘要：There are two important components in a meta-heuristic algorithm, namely, Diversification and Intensification. These components can be fine-tuned for the optimized execution of a meta-heuristic search algorithm. The firefly algorithm (FA) is the latest in a series of meta-heuristic algorithms. Although the FA has proven to be efficient in local searches, there are times when it might get trapped in several local optima, as a result of which it is unable to efficiently conduct a complete search. In the pursuit of global space scaling, this algorithm needs to generate different solutions leveraging on diversification. The function of updating the effectiveness of diversification in a search algorithm can be performed by elitism operators. In this study, a strategy was proposed to upgrade the FA concerning static issues. The methodology involved the hybridization of elitism with the standard firefly algorithm, and this modified version was known as the AFA. Also, another distribution was introduced to revamp the entire search process in the FA using a t-way test generation (t referring to the strength of the interaction). The experimental results demonstrated that the Elitism firefly algorithm (eFA) had better performance than the standard FA and also than the other up to date algorithms in terms of robustness and the convergence speed.
关键词：Firefly Algorithm; Elitism; T-way testing; Intensification; And Diversification.
