Scholarlink

<p>Mapping optimal areas for finding geomorphosites in large areas is a complex task influenced by various factors. To address this challenge, the present study assesses the effectiveness of three widely used machine learning classifiers in identifying and mapping potential areas for geomorphosite, a crucial factor for attracting and enhancing geotourism in Ziz, southeast Morocco. Initially, a comprehensive inventory of 120 geomorphosites was conducted in the study area. At each site, precise measurements of three topographical parameters were taken. Following this, three machine learning algorithms, namely Random Forest, Multi-Layer Perceptron, and M5 Prime, were utilized to create predictive models. Regarding the model performance, the Multi-Layer Perceptron model achieved the highest performance with an area under the curve of 0.91, followed by the M5 Prime model with 0.77. These models identified highly favorable areas, which accounted for approximately 60% and 42% of the study area according to Multi-Layer Perceptron and M5 Prime, respectively. These areas were predominantly located in the western region, characterized by mountainous terrain with relatively shorter slope lengths and altitudes ranging between 2500&#xa0;m and 3500&#xa0;m. This research serves as a valuable roadmap for decision-makers, offering guidance on how to improve the likelihood of discovering geomorphosites while minimizing costs and reducing the time required for exploration.</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 Geotourism in Southeastern Morocco through Machine Learning-Based Geomorphosite Identification

  • 期刊论文

  • 发表日期：

    2025年2月14日

查看全文
  • Mohamed Manaouch,
  • Lahbib Naimi,
  • Mbarek Haynou,
  • Mohamed Aghad,
  • Mohamed Sadiki,
  • Quoc Bao Pham,
  • Abdeslam Jakimi

摘要

Mapping optimal areas for finding geomorphosites in large areas is a complex task influenced by various factors. To address this challenge, the present study assesses the effectiveness of three widely used machine learning classifiers in identifying and mapping potential areas for geomorphosite, a crucial factor for attracting and enhancing geotourism in Ziz, southeast Morocco. Initially, a comprehensive inventory of 120 geomorphosites was conducted in the study area. At each site, precise measurements of three topographical parameters were taken. Following this, three machine learning algorithms, namely Random Forest, Multi-Layer Perceptron, and M5 Prime, were utilized to create predictive models. Regarding the model performance, the Multi-Layer Perceptron model achieved the highest performance with an area under the curve of 0.91, followed by the M5 Prime model with 0.77. These models identified highly favorable areas, which accounted for approximately 60% and 42% of the study area according to Multi-Layer Perceptron and M5 Prime, respectively. These areas were predominantly located in the western region, characterized by mountainous terrain with relatively shorter slope lengths and altitudes ranging between 2500 m and 3500 m. This research serves as a valuable roadmap for decision-makers, offering guidance on how to improve the likelihood of discovering geomorphosites while minimizing costs and reducing the time required for exploration.