Scholarlink

<p>The China–Pakistan Economic Corridor (CPEC) is one of the world’s mega-projects and is composed of roads, railways, oil and gas pipelines, fiber-optic cables, and power lines, stretching from China’s Kashgar to Pakistan’s Gwadar Port. This study was conducted to evaluate the occurrence and distribution of landslides along the Karakoram Highway (KKH). KKH is one of the most dangerous highways in the world due to its location in the high mountain ranges. Therefore, this is the most critical part of the corridor between China and Pakistan to study natural disasters such as landslides. Quantitative methods, including the analytic hierarchy process (AHP)–based weighted overlay model (WOM) and weighted information model (WIM), were used alongside machine learning (ML) models such as extreme gradient boost (XGBoost or XGB) and random forest (RF). The accuracy of these models, defined as the proportion of correct predictions among the total number of cases, was assessed using the receiver operating characteristic (ROC) curve and additional statistical performance measures. The findings indicated that, among the quantitative methods, the WIM demonstrated higher accuracy (0.886) than the AHP model (0.817). In the case of ML models, the XGBoost model achieved higher accuracy (0.898) than the AHP model (0.875), indicating superior performance in assessing susceptibility. The results provide the scientific foundation for major project design, traffic route selection, urban planning, resource creation, and other aspects of the CPEC that will benefit regional people’s livelihood protection and long-term development.</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

Landslide risk assessment on the China–Pakistan Economic Corridor (CPEC): a comparative study of quantitative and machine learning approaches

  • 期刊论文

  • 发表日期：

    2025年5月5日

查看全文
  • Aboubakar Siddique,
  • Tan Qulin,
  • Hilal Ahmad,
  • Umair Rasool,
  • Zhuoying Tan,
  • Meer Muhammad Sajjad,
  • Farhan Iftikharf,
  • Mansour Shrahili,
  • Shafeeq ur Rahman

摘要

The China–Pakistan Economic Corridor (CPEC) is one of the world’s mega-projects and is composed of roads, railways, oil and gas pipelines, fiber-optic cables, and power lines, stretching from China’s Kashgar to Pakistan’s Gwadar Port. This study was conducted to evaluate the occurrence and distribution of landslides along the Karakoram Highway (KKH). KKH is one of the most dangerous highways in the world due to its location in the high mountain ranges. Therefore, this is the most critical part of the corridor between China and Pakistan to study natural disasters such as landslides. Quantitative methods, including the analytic hierarchy process (AHP)–based weighted overlay model (WOM) and weighted information model (WIM), were used alongside machine learning (ML) models such as extreme gradient boost (XGBoost or XGB) and random forest (RF). The accuracy of these models, defined as the proportion of correct predictions among the total number of cases, was assessed using the receiver operating characteristic (ROC) curve and additional statistical performance measures. The findings indicated that, among the quantitative methods, the WIM demonstrated higher accuracy (0.886) than the AHP model (0.817). In the case of ML models, the XGBoost model achieved higher accuracy (0.898) than the AHP model (0.875), indicating superior performance in assessing susceptibility. The results provide the scientific foundation for major project design, traffic route selection, urban planning, resource creation, and other aspects of the CPEC that will benefit regional people’s livelihood protection and long-term development.