Scholarlink

<p>This paper presents a direct-plug-in bandwidth selector for local quadratic regression and local cubic regression, leveraging existing theoretical frameworks. Through extensive simulation studies, the performance of the proposed selector is evaluated using the Mean Squared Error (MSE) and Mean Absolute Error (MAE) criteria, in comparison with established methods. Additionally, empirical coverage of confidence intervals is analyzed to further assess its effectiveness. Practical applications of the methods are illustrated using wildfire rate of spread data.</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

A comparison of bandwidth selectors for moderate degree local polynomial regression

  • 期刊论文

  • 发表日期：

    2025年4月2日

查看全文
  • Dongying Wang,
  • W. John Braun

摘要

This paper presents a direct-plug-in bandwidth selector for local quadratic regression and local cubic regression, leveraging existing theoretical frameworks. Through extensive simulation studies, the performance of the proposed selector is evaluated using the Mean Squared Error (MSE) and Mean Absolute Error (MAE) criteria, in comparison with established methods. Additionally, empirical coverage of confidence intervals is analyzed to further assess its effectiveness. Practical applications of the methods are illustrated using wildfire rate of spread data.