Scholarlink

<p>This article introduces the Salmon Magnetic Odor Algorithm (SMOA), an optimization meta-heuristic based on the behavior of salmon which return to their original rivers for spawning, informing how algorithms may be applied in a way that is more natural and efficient. It has three mechanisms that work in a parallel manner: 1) – an exploitation strategy based on the attraction of a magnetic field; 2) Local Search for Global Exploration Based on Scent; 3) dynamic parameter adaptation exponential decay, sinusoidal and diversity based adaptations.</p><p>Extensive experiments demonstrate that SMOA consistently outperforms recent state-of-the-art optimizers, including BinDMO (2024), WPO (2023), SGA (2023), RSA (2022), and ZOA (2022). SMOA achieves optimal or near-optimal results in most performance tests, with top values such as 5.32e + 02 for F1, 2.00e + 02 for F2, and 1.36e + 03 for F9, often surpassing the accuracy of its competitors by several orders of magnitude. Furthermore, SMOA stands out for its unparalleled reliability, characterized by a low margin of error and remarkably stable results over several iterations. These results highlight the efficiency, robustness, and potential of SMOA as a particularly effective optimization algorithm for complex numerical problems. With the ability to adaptively control and balance exploration–exploitation, SMOA is indeed a strong and versatile approach for tackling complex high-dimensional optimization problems. While other optimizers sometimes fail (meaning they either cannot extract good results for certain footsteps or their error bars may be inconsiderable at best), SMOA always wins with its great ability to control randomness: the margin of error is practically zero in the vast majority of cases.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Salmon magnetic odor optimization algorithm (SMOA)

  • 期刊论文

  • 发表日期：

    2025年10月3日

查看全文
  • Hassan Afifi,
  • Idriss Chana,
  • Kaoutar Baibai,
  • Aziz Bouazi

摘要

This article introduces the Salmon Magnetic Odor Algorithm (SMOA), an optimization meta-heuristic based on the behavior of salmon which return to their original rivers for spawning, informing how algorithms may be applied in a way that is more natural and efficient. It has three mechanisms that work in a parallel manner: 1) – an exploitation strategy based on the attraction of a magnetic field; 2) Local Search for Global Exploration Based on Scent; 3) dynamic parameter adaptation exponential decay, sinusoidal and diversity based adaptations.

Extensive experiments demonstrate that SMOA consistently outperforms recent state-of-the-art optimizers, including BinDMO (2024), WPO (2023), SGA (2023), RSA (2022), and ZOA (2022). SMOA achieves optimal or near-optimal results in most performance tests, with top values such as 5.32e + 02 for F1, 2.00e + 02 for F2, and 1.36e + 03 for F9, often surpassing the accuracy of its competitors by several orders of magnitude. Furthermore, SMOA stands out for its unparalleled reliability, characterized by a low margin of error and remarkably stable results over several iterations. These results highlight the efficiency, robustness, and potential of SMOA as a particularly effective optimization algorithm for complex numerical problems. With the ability to adaptively control and balance exploration–exploitation, SMOA is indeed a strong and versatile approach for tackling complex high-dimensional optimization problems. While other optimizers sometimes fail (meaning they either cannot extract good results for certain footsteps or their error bars may be inconsiderable at best), SMOA always wins with its great ability to control randomness: the margin of error is practically zero in the vast majority of cases.