Scholarlink

<p>Control charts based on regression models are important to control quality and/or process, or to improve overall healthcare in both industrial and medical fields. Conventional control charts have difficulty dealing with outliers that diminish their sensitivity to significant changes. Proposing a new hybridization of the Hampel filter for outlier detection and robust regression methods, this study aims to enhance the performance of control charts that monitor standardized regression slopes and mean squared errors. Because it minimizes the effect of outlier data points, the proposed method can produce control charts with tighter limits and improved sensitivity for detecting process variations. It is validated via extensive simulations and real medical data, positively comparing existing approaches in its ability to capture abnormal variations with fewer false alarms. The study demonstrates the applicability of the technique as a useful and valid approach for improving quality control in complicated industrial and industrial settings.</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

Novel Hampel Filtering and Robust Regression in Control Chart Applications for Regression Process Monitoring

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年10月10日

查看全文
  • Taha Hussein Ali,
  • Hutheyfa Hazem Taha,
  • Bekhal Samad Sedeeq,
  • Heyam A. A. Hayawi

摘要

Control charts based on regression models are important to control quality and/or process, or to improve overall healthcare in both industrial and medical fields. Conventional control charts have difficulty dealing with outliers that diminish their sensitivity to significant changes. Proposing a new hybridization of the Hampel filter for outlier detection and robust regression methods, this study aims to enhance the performance of control charts that monitor standardized regression slopes and mean squared errors. Because it minimizes the effect of outlier data points, the proposed method can produce control charts with tighter limits and improved sensitivity for detecting process variations. It is validated via extensive simulations and real medical data, positively comparing existing approaches in its ability to capture abnormal variations with fewer false alarms. The study demonstrates the applicability of the technique as a useful and valid approach for improving quality control in complicated industrial and industrial settings.