Scholarlink

<p>The study offers a significant advancement in the field of groundwater potential (GWP) mapping, specifically for arid regions, through the innovative integration of clustering algorithms with the Dempster–Shafer theory (DST). The proposed methodology optimizes GWP assessments and enhances the precision and reliability of groundwater management decisions in regions like Khor and Biabanak, Iran. By integrating multiple clustering techniques—K-means, expectation–maximization, and agglomerative hierarchical clustering—within the DST framework, the model is able to reduce uncertainty and deliver high-accuracy maps for groundwater resource management. Through data-driven optimization using Silhouette and Calinski–Harabasz indices, the study identifies the most reliable clustering configuration, improving the precision of traditional classification techniques such as natural breaks and equal intervals. The model's ability to classify areas based on geological, topographic, and soil data enables targeted interventions in areas of high groundwater potential, supporting more sustainable water management practices. With validation accuracies of 86% for training and 84% for testing, the approach demonstrates high efficacy in predicting GWP, making it an invaluable tool for regional water resource management.</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

Integrating Dempster–Shafer theory and clustering algorithms for enhanced groundwater potential assessment

  • 期刊论文

  • 发表日期：

    2025年8月5日

查看全文
  • Ali Azizi,
  • Parham Pahlavani,
  • Mohammad Nakhaei

摘要

The study offers a significant advancement in the field of groundwater potential (GWP) mapping, specifically for arid regions, through the innovative integration of clustering algorithms with the Dempster–Shafer theory (DST). The proposed methodology optimizes GWP assessments and enhances the precision and reliability of groundwater management decisions in regions like Khor and Biabanak, Iran. By integrating multiple clustering techniques—K-means, expectation–maximization, and agglomerative hierarchical clustering—within the DST framework, the model is able to reduce uncertainty and deliver high-accuracy maps for groundwater resource management. Through data-driven optimization using Silhouette and Calinski–Harabasz indices, the study identifies the most reliable clustering configuration, improving the precision of traditional classification techniques such as natural breaks and equal intervals. The model's ability to classify areas based on geological, topographic, and soil data enables targeted interventions in areas of high groundwater potential, supporting more sustainable water management practices. With validation accuracies of 86% for training and 84% for testing, the approach demonstrates high efficacy in predicting GWP, making it an invaluable tool for regional water resource management.