Scholarlink

<p>As the blue engine driving global economic growth, the marine economy is playing an increasingly prominent role in national strategies. However, due to the limited sample size and strong nonlinearity of its development data, traditional forecasting methods often suffer from deficiencies in adaptability and accuracy. To address these challenges, an adaptive nonlinear grey forecasting model by introducing both linear and nonlinear driving terms is proposed in this study, and by jointly optimizing the background value, accumulation order, and time power parameters to enhance model performance. To further improve parameter stability, a hybrid optimization framework combining intelligent optimization algorithm with hierarchical clustering is designed, enabling the selection of the optimal parameter configuration through clustering of multiple optimization results. In an empirical study on marine economic trend forecasting, the proposed approach substantially improves prediction accuracy, demonstrating its applicability and potential for broader use in modeling complex systems.</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

Parameter optimization of adaptive nonlinear grey prediction model and its prediction of marine economic trend

  • 期刊论文

  • 发表日期：

    2026年2月12日

查看全文
  • Yimei Jin,
  • Xiaoyi Gou,
  • Yuhan Xie

摘要

As the blue engine driving global economic growth, the marine economy is playing an increasingly prominent role in national strategies. However, due to the limited sample size and strong nonlinearity of its development data, traditional forecasting methods often suffer from deficiencies in adaptability and accuracy. To address these challenges, an adaptive nonlinear grey forecasting model by introducing both linear and nonlinear driving terms is proposed in this study, and by jointly optimizing the background value, accumulation order, and time power parameters to enhance model performance. To further improve parameter stability, a hybrid optimization framework combining intelligent optimization algorithm with hierarchical clustering is designed, enabling the selection of the optimal parameter configuration through clustering of multiple optimization results. In an empirical study on marine economic trend forecasting, the proposed approach substantially improves prediction accuracy, demonstrating its applicability and potential for broader use in modeling complex systems.