Scholarlink

<p>In 2023, the Crayfish Optimization Algorithm (COA) was introduced as a type of meta-heuristic optimization algorithm, inspired by crayfish behavior. While COA exhibits strong optimization performance, it encounters challenges such as slow convergence and susceptibility to local optima. To address these issues, this paper introduces the Multi-strategy improvement of crayfish optimization algorithm to solve high-dimensional feature selection (MICOA), building upon COA's foundation. Specifically, MICOA enhances the original algorithm by introducing the cave selection strategy, mitigating the tendency to converge to local optima. Additionally, it incorporates a food attraction strategy based on crayfish behavior, fostering a more balanced exploration–exploitation trade-off and enhancing convergence speed. The paper verifies MICOA's performance against the original COA and six comparison algorithms. Results on CEC2020 and CEC2014 demonstrate MICOA's enhanced ability to overcome local optima and converge faster. In terms of data analysis, the enhanced MICOA demonstrates an average 20% enhancement in convergence capability compared to the original COA. Furthermore, when contrasted with alternative algorithms, MICOA showcases a maximum improvement of 75%. Applied to high-dimensional feature selection datasets, MICOA outperforms the original algorithm and six other methods, achieving the highest accuracy on most datasets. MICOA provides a significant 70% improvement in accuracy compared to the original COA for feature selection applications. In addition, MICOA has a significant advantage over the prevalent feature selection algorithms currently in use. These results highlight the superiority of MICOA over existing methods.</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

Multi-strategy improvement of crayfish optimization algorithm to solve high-dimensional feature selection

  • 期刊论文

  • 发表日期：

    2025年4月28日

查看全文
  • Xiaoming Shi,
  • Heming Jia,
  • Honghua Rao,
  • Fangkai You,
  • Laith Abualigah

摘要

In 2023, the Crayfish Optimization Algorithm (COA) was introduced as a type of meta-heuristic optimization algorithm, inspired by crayfish behavior. While COA exhibits strong optimization performance, it encounters challenges such as slow convergence and susceptibility to local optima. To address these issues, this paper introduces the Multi-strategy improvement of crayfish optimization algorithm to solve high-dimensional feature selection (MICOA), building upon COA's foundation. Specifically, MICOA enhances the original algorithm by introducing the cave selection strategy, mitigating the tendency to converge to local optima. Additionally, it incorporates a food attraction strategy based on crayfish behavior, fostering a more balanced exploration–exploitation trade-off and enhancing convergence speed. The paper verifies MICOA's performance against the original COA and six comparison algorithms. Results on CEC2020 and CEC2014 demonstrate MICOA's enhanced ability to overcome local optima and converge faster. In terms of data analysis, the enhanced MICOA demonstrates an average 20% enhancement in convergence capability compared to the original COA. Furthermore, when contrasted with alternative algorithms, MICOA showcases a maximum improvement of 75%. Applied to high-dimensional feature selection datasets, MICOA outperforms the original algorithm and six other methods, achieving the highest accuracy on most datasets. MICOA provides a significant 70% improvement in accuracy compared to the original COA for feature selection applications. In addition, MICOA has a significant advantage over the prevalent feature selection algorithms currently in use. These results highlight the superiority of MICOA over existing methods.