Comparative study of different machine learning models in landslide susceptibility assessment: A case study of Conghua District, Guangzhou, China

IF 4.6 3区地球科学 Q1 GEOSCIENCES, MULTIDISCIPLINARY

China Geology Pub Date : 2024-01-01 DOI:10.31035/cg2023056

Ao Zhang , Xin-wen Zhao , Xing-yuezi Zhao , Xiao-zhan Zheng , Min Zeng , Xuan Huang , Pan Wu , Tuo Jiang , Shi-chang Wang , Jun He , Yi-yong Li

{"title":"Comparative study of different machine learning models in landslide susceptibility assessment: A case study of Conghua District, Guangzhou, China","authors":"Ao Zhang , Xin-wen Zhao , Xing-yuezi Zhao , Xiao-zhan Zheng , Min Zeng , Xuan Huang , Pan Wu , Tuo Jiang , Shi-chang Wang , Jun He , Yi-yong Li","doi":"10.31035/cg2023056","DOIUrl":null,"url":null,"abstract":"<div><p>Machine learning is currently one of the research hotspots in the field of landslide prediction. To clarify and evaluate the differences in characteristics and prediction effects of different machine learning models, Conghua District, which is the most prone to landslide disasters in Guangzhou, was selected for landslide susceptibility evaluation. The evaluation factors were selected by using correlation analysis and variance expansion factor method. Applying four machine learning methods namely Logistic Regression (LR), Random Forest (RF), Support Vector Machines (SVM), and Extreme Gradient Boosting (XGB), landslide models were constructed. Comparative analysis and evaluation of the model were conducted through statistical indices and receiver operating characteristic (ROC) curves. The results showed that LR, RF, SVM, and XGB models have good predictive performance for landslide susceptibility, with the area under curve (AUC) values of 0.752, 0.965, 0.996, and 0.998, respectively. XGB model had the highest predictive ability, followed by RF model, SVM model, and LR model. The frequency ratio (FR) accuracy of LR, RF, SVM, and XGB models was 0.775, 0.842, 0.759, and 0.822, respectively. RF and XGB models were superior to LR and SVM models, indicating that the integrated algorithm has better predictive ability than a single classification algorithm in regional landslide classification problems.</p><p>©2024 China Geology Editorial Office.</p></div>","PeriodicalId":45329,"journal":{"name":"China Geology","volume":null,"pages":null},"PeriodicalIF":4.6000,"publicationDate":"2024-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2096519224000594/pdfft?md5=7dfe5a4b1b90b0a27e8db257b6c84e46&pid=1-s2.0-S2096519224000594-main.pdf","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"China Geology","FirstCategoryId":"89","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S2096519224000594","RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"GEOSCIENCES, MULTIDISCIPLINARY","Score":null,"Total":0}

引用次数: 0

Abstract

Machine learning is currently one of the research hotspots in the field of landslide prediction. To clarify and evaluate the differences in characteristics and prediction effects of different machine learning models, Conghua District, which is the most prone to landslide disasters in Guangzhou, was selected for landslide susceptibility evaluation. The evaluation factors were selected by using correlation analysis and variance expansion factor method. Applying four machine learning methods namely Logistic Regression (LR), Random Forest (RF), Support Vector Machines (SVM), and Extreme Gradient Boosting (XGB), landslide models were constructed. Comparative analysis and evaluation of the model were conducted through statistical indices and receiver operating characteristic (ROC) curves. The results showed that LR, RF, SVM, and XGB models have good predictive performance for landslide susceptibility, with the area under curve (AUC) values of 0.752, 0.965, 0.996, and 0.998, respectively. XGB model had the highest predictive ability, followed by RF model, SVM model, and LR model. The frequency ratio (FR) accuracy of LR, RF, SVM, and XGB models was 0.775, 0.842, 0.759, and 0.822, respectively. RF and XGB models were superior to LR and SVM models, indicating that the integrated algorithm has better predictive ability than a single classification algorithm in regional landslide classification problems.

查看原文本刊更多论文

不同机器学习模型在滑坡易感性评估中的比较研究：中国广州市从化区案例研究

机器学习是当前滑坡预测领域的研究热点之一。为明确和评价不同机器学习模型的特点和预测效果的差异，选取广州市滑坡灾害最易发生的从化区进行滑坡易感性评价。评价因子的选择采用相关分析法和方差扩展因子法。应用四种机器学习方法，即逻辑回归（LR）、随机森林（RF）、支持向量机（SVM）和极梯度提升（XGB），构建了滑坡模型。通过统计指数和接收者工作特征曲线（ROC）对模型进行了比较分析和评估。结果表明，LR、RF、SVM 和 XGB 模型对滑坡易感性具有良好的预测性能，其曲线下面积（AUC）值分别为 0.752、0.965、0.996 和 0.998。XGB 模型的预测能力最高，其次是 RF 模型、SVM 模型和 LR 模型。LR、RF、SVM 和 XGB 模型的频率比（FR）准确率分别为 0.775、0.842、0.759 和 0.822。RF和XGB模型优于LR和SVM模型，表明在区域滑坡分类问题上，综合算法比单一分类算法具有更好的预测能力。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊