Scholarlink

<p>This paper introduces the Side Information Boosted Symbolic Regression (SIBSR) model, an enhanced approach in symbolic regression aimed at improving data analysis. SIBSR integrates side information to increase the accuracy and efficiency of modeling complex data relationships. In addition, we introduce the Side Information Generator, a complementary tool designed to assist in generating a range of potential side information options. This enables users to select the most effective side information for specific tasks, thereby enhancing practical utility. Our experimental findings demonstrate the efficacy of SIBSR in standard symbolic regression tasks and its practical application in economic contexts, notably in formulating Nash Equilibrium expressions in Game Theory. These results underscore SIBSR’s potential in advancing the field of data analysis. The source codes are available at: <a href="https://github.com/dkflame/SIBSR">https://github.com/dkflame/SIBSR</a>.</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

Enhancing symbolic regression with side information for data analysis

  • 期刊论文

  • 发表日期：

    2025年1月9日

查看全文
  • Xiangwu Zuo,
  • Anxiao Jiang

摘要

This paper introduces the Side Information Boosted Symbolic Regression (SIBSR) model, an enhanced approach in symbolic regression aimed at improving data analysis. SIBSR integrates side information to increase the accuracy and efficiency of modeling complex data relationships. In addition, we introduce the Side Information Generator, a complementary tool designed to assist in generating a range of potential side information options. This enables users to select the most effective side information for specific tasks, thereby enhancing practical utility. Our experimental findings demonstrate the efficacy of SIBSR in standard symbolic regression tasks and its practical application in economic contexts, notably in formulating Nash Equilibrium expressions in Game Theory. These results underscore SIBSR’s potential in advancing the field of data analysis. The source codes are available at: https://github.com/dkflame/SIBSR.