Improved Manta Ray Foraging Optimizer-based SVM for Feature Selection Problems: A Medical Case Study

Got, Adel; Zouache, Djaafar; Moussaoui, Abdelouahab; Abualigah, Laith; Alsayat, Ahmed

doi:10.1007/s42235-023-00436-9

Cited by 4 publications

(1 citation statement)

References 64 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Such algorithms do not depend on the specific form of the problem but rather guide the search process by simulating phenomena in nature, behaviors of organisms, physical principles, and social laws, etc., thus demonstrating excellent adaptability and efficiency in numerous application fields. With the continuous development of metaheuristic algorithms, these algorithms play a crucial role in a variety of fields, such as path planning [2,3], image segmentation [4,5], feature selection [6,7], neural network hyperparameter optimization [8,9], task allocation [10,11], supply chain management [12,13], waste collection [14], wireless sensor optimization problems [15,16], and antenna array synthesis issues [17,18]. And, they show great potential in promoting the development of engineering technology, improving productivity, and solving multi-objective optimization problems [19,20].…”

Section: Introductionmentioning

confidence: 99%

Improved Multi-Strategy Sand Cat Swarm Optimization for Solving Global Optimization

Zhang,

He,

Wang

et al. 2024

Biomimetics

View full text Add to dashboard Cite

The sand cat swarm optimization algorithm (SCSO) is a novel metaheuristic algorithm that has been proposed in recent years. The algorithm optimizes the search ability of individuals by mimicking the hunting behavior of sand cat groups in nature, thereby achieving robust optimization performance. It is characterized by few control parameters and simple operation. However, due to the lack of population diversity, SCSO is less efficient in solving complex problems and is prone to fall into local optimization. To address these shortcomings and refine the algorithm’s efficacy, an improved multi-strategy sand cat optimization algorithm (IMSCSO) is proposed in this paper. In IMSCSO, a roulette fitness–distance balancing strategy is used to select codes to replace random agents in the exploration phase and enhance the convergence performance of the algorithm. To bolster population diversity, a novel population perturbation strategy is introduced, aiming to facilitate the algorithm’s escape from local optima. Finally, a best–worst perturbation strategy is developed. The approach not only maintains diversity throughout the optimization process but also enhances the algorithm’s exploitation capabilities. To evaluate the performance of the proposed IMSCSO, we conducted experiments in the CEC 2017 test suite and compared IMSCSO with seven other algorithms. The results show that the IMSCSO proposed in this paper has better optimization performance.

show abstract