Greylag Goose Optimization: Nature-inspired optimization algorithm

• Published in: Expert systems with applications Vol. 238; p. 122147
• Main Authors: El-kenawy, El-Sayed M., Khodadadi, Nima, Mirjalili, Seyedali, Abdelhamid, Abdelaziz A., Eid, Marwa M., Ibrahim, Abdelhameed
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 15.03.2024
• Subjects: Algorithm; Engineering optimization; Meta-heuristic; Swarm-based algorithms; Swarm-based algorithms; Meta-heuristic; Algorithm; Engineering optimization
• ISSN: 0957-4174; 1873-6793
• DOI: 10.1016/j.eswa.2023.122147

Nature-inspired metaheuristic approaches draw their core idea from biological evolution in order to create new and powerful competing algorithms. Such algorithms can be divided into evolution-based and swarm-based algorithms. This paper proposed a new nature-inspired optimizer called the Greylag Goose Optimization (GGO) algorithm. The proposed algorithm (GGO) belongs to the class of swarm-based algorithms and is inspired by the Greylag Goose. Geese are excellent flyers and during their seasonal migrations, they fly in a group and can cover thousands of kilometers in a single flight. While flying, a group of geese forms themselves as a “V” configuration. In this way, the geese in the front can minimize the air resistance of the ones in the back. This allows the geese to fly around 70% farther as a group than they could individually. The GGO algorithm is first validated by being applied to nineteen datasets retrieved from the UCI Machine Learning Repository. Each dataset contains a varied amount of characteristics, instances, and classes that are used to choose features. After that, it is put to use in the process of solving a number of engineering benchmark functions and case studies. Several case studies are solved using the proposed algorithm too, including the pressure vessel design and the tension/compression spring design. The findings demonstrate that the GGO method outperforms numerous other comparative optimization algorithms and delivers superior accuracy compared to other algorithms. The results of the statistical analysis tests, such as Wilcoxon’s rank-sum and one-way analysis of variance (ANOVA), demonstrate that the GGO algorithm achieves superior results. •Greylag Goose Optimization (GGO) algorithm inspired by nature is proposed.•A binary version of the GGO algorithm is designed for feature selection.•The GGO algorithm is tested on 19 datasets from UCI Machine Learning Repository.•The GGO algorithm is applied to engineering problems and benchmark functions.•The statistical significance of the GGO algorithm has been evaluated.