连续变量因子分级和机器学习模型对滑坡易发性评价精度的影响

郭飞, 吴迪, 葛民荣, 董进龙, 房浩, 田东方

郭飞, 吴迪, 葛民荣, 董进龙, 房浩, 田东方. 连续变量因子分级和机器学习模型对滑坡易发性评价精度的影响[J]. 武汉大学学报 ( 信息科学版). DOI: 10.13203/j.whugis20230413
引用本文: 郭飞, 吴迪, 葛民荣, 董进龙, 房浩, 田东方. 连续变量因子分级和机器学习模型对滑坡易发性评价精度的影响[J]. 武汉大学学报 ( 信息科学版). DOI: 10.13203/j.whugis20230413
GUO Fei, WU Di, GE Minrong, DONG Jinlong, FANG Hao, TIAN Dongfang. The Influence of Continuous Variable Factor Classification and Machine Learning Model on the Accuracy of Landslide Susceptibility Evaluation[J]. Geomatics and Information Science of Wuhan University. DOI: 10.13203/j.whugis20230413
Citation: GUO Fei, WU Di, GE Minrong, DONG Jinlong, FANG Hao, TIAN Dongfang. The Influence of Continuous Variable Factor Classification and Machine Learning Model on the Accuracy of Landslide Susceptibility Evaluation[J]. Geomatics and Information Science of Wuhan University. DOI: 10.13203/j.whugis20230413

连续变量因子分级和机器学习模型对滑坡易发性评价精度的影响

基金项目: 

国家自然科学基金(42107489)

国家重点研发计划(2021YFC3001901、 2022YFC3005603)

湖北巴东地质灾害国家野外科学观测研究站开放基金(BNORSG202304)

三峡库区地质灾害教育部重点实验室开放基金(2022KDZ14)

湖北省自然科学基金(2022CFB557)

土木工程防灾减灾湖北省引智创新示范基地项目(2021EJD026)。

详细信息
    作者简介:

    郭飞,博士,副教授,主要从事区域地质灾害风险评估研究。ybbnui.2008@163.com

    通讯作者:

    田东方,博士, 教授。tdf_2005@163.com

The Influence of Continuous Variable Factor Classification and Machine Learning Model on the Accuracy of Landslide Susceptibility Evaluation

  • 摘要: 滑坡易发性评价建模中环境因子分级区间和机器学习模型对建模结果的影响不容忽视。为探究这两种因素对滑坡易发性评价结果的影响规律, 基于主客观赋权法通过对环境因子进行赋权以构建评价指标体系, 再利用地理探测器探究不同连续变量因子分级对滑坡易发性评价结果精度的影响规律,进而分别采用随机森林模型、 梯度极限提升模型和遗传算法优化的神经网络模型开展滑坡易发性研究。结果表明: 1) 通过地理探测器得到的与灾害关联度最高的分区组合计算出的最大 AUC 值为 0.886,说明该方法可以得到最优的分级区间,且能有效提高易发性评价结果的精度; 2) 在易发性评价结果中, 随机森林模型最优, 较梯度极限提升模型和遗传算法优化的神经网络模型精度分别提高了 9.7%和 9.6%。基于地理探测器的环境因子最优分级区间是合理的, 且随机森林模型作为滑坡易发性评价模型是高效准确的。
    Abstract: Objectives: The influence of environmental factor classification interval and machine learning model on modeling results in landslide susceptibility evaluation modeling cannot be ignored. In order to explore the influence of these two factors on the evaluation results of landslide susceptibility, Methods: the evaluation index system is constructed by weighting the environmental factors based on the subjective and objective weighting method, and then the influence of different continuous variable factor classification on the accuracy of landslide susceptibility evaluation results is explored by using the GeoDetector. Then, the random forest model, the gradient limit lifting model and the neural network model optimized by genetic algorithm are used to study the landslide susceptibility. Results: The results show that: 1) The maximum AUC value calculated by the partition combination with the highest correlation degree with the disaster obtained by the GeoDetector is 0.886, indicating that the method can obtain the optimal classification interval and can effectively improve the accuracy of the susceptibility evaluation results. 2) In the susceptibility evaluation results, the random forest model is the best, which is 9.7% and 9.6% higher than the gradient limit lifting model and the neural network model optimized by genetic algorithm. Conclusions: The optimal classification interval of environmental factors based on GeoDetector is reasonable, and the random forest model is efficient and accurate as a landslide susceptibility evaluation model.
  • [1]

    PETLEY D. Global patterns of loss of life from landslides[J]. Geology, 2012, 40(10): 927-930.

    [2] XU Shenghua, LIU Jiping, WANG Xianghong, ZHANG Yu, LIN Rongfu, ZHANG Meng, LIU Mengmeng, JIANG Tao. Landslide Susceptibility Assessment Method Incorporating Index of Entropy Based on Support Vector Machine: A Case Study of Shaanxi Province[J]. Geomatics and Information Science of Wuhan University, 2020, 45(8): 1214-1222. DOI: 10.13203/j.whugis20200109(徐胜华, 刘纪平, 王想红, 张玉, 林荣福, 张蒙, 刘猛猛, 姜涛. 熵指数融入支持向量机的滑坡灾害易发性评价方法—以陕西省为例[J]. 武汉大学学报(信息科学版), 2020, 45(8): 1214-1222. DOI: 10.13203/j.whugis20200109)
    [3]

    CHEN Y, DONG J, GUO F, et al. Review of landslide susceptibility assessment based on knowledge mapping[J]. Stochastic Environmental Research and Risk Assessment, 2022, 36(9): 2399-2417.

    [4] YAN Jusheng, TAN Jianmin. Landslide susceptibility evaluation based on different factor classification methods-taking Yuan'an County, Hubei Province as an example [J]. Chinese Journal of Geological Disasters and Prevention, 2019(1): 52-60.(闫举生, 谭建民. 基于不同因子分级法的滑坡易发性评价—以湖北远安县为例[J]. 中国地质灾害与防治学报, 2019(1): 52-60.)
    [5] LI Ping, YE Hui, TAN Shucheng. Evaluation of geological disaster susceptibility in Yongde County based on analytic hierarchy process [J]. Soil and Water Conservation Research, 2021,28(5): 394-399.(李萍, 叶辉, 谈树成. 基于层次分析法的永德县地质灾害易发性评价[J]. 水土保持研究, 2021, 28(5): 394-399.)
    [6] LIAN Zhipeng, XU Yong, FU Sheng, et al. Multi-model fusion method was used to evaluate the susceptibility of landslide disasters: a case study of Wufeng County, Hubei Province [J]. Geological Science and Technology Bulletin, 2020,39(3): 178-186.(连志鹏, 徐勇, 付圣, 等. 采用多模型融合方法评价滑坡灾害易发性: 以湖北省五峰县为例[J]. 地质科技通报, 2020, 39(3): 178-186.)
    [7] WU Chenwen, LIANG Jinghan, WANG Wei, et al. Random forest algorithm based on recursive feature elimination method [J]. Statistics and decision-making, 2017(21): 60-63.(吴辰文, 梁靖涵, 王伟, 等. 基于递归特征消除方法的随机森林算法[J]. 统计与决策, 2017(21): 60-63.)
    [8] YOU Shibing, YAN Yan. Stepwise Regression Analysis and Its Application [J]. Statistics and Decision Making, 2017(14): 31- 35.(游士兵, 严研. 逐步回归分析法及其应用[J]. 统计与决策, 2017(14): 31-35.)
    [9]

    LUO W, LIU C C. Innovative landslide susceptibility mapping supported by geomorphon and geographical detector methods[J]. Landslides, 2018, 15: 465-474.

    [10] LIU Ting, TAN Jianmin, GUO Fei, et al. Study on the weight correction method in the evaluation of landslide susceptibility under artificial cutting slope - A case study of Shadi Town, Ganzhou City [J]. Journal of Natural Disasters, 2021, 30(5): 217-225.(刘婷, 谭建民, 郭飞, 等. 人工切坡下滑坡易发性评价中权重修正方法研究—以赣州市沙地镇为例[J]. 自然灾害学报, 2021, 30(5): 217-225.)
    [11]

    CHANG M, ZHOU Y, ZHOU C, et al. Coseismic landslides induced by the 2018 M w 6.6Iburi, Japan, Earthquake: spatial distribution, key factors weight, and susceptibility regionalization[J]. Landslides, 2021, 18: 755-772.

    [12]

    HUANG F, YE Z, JIANG S H, et al. Uncertainty study of landslide susceptibility prediction considering the different attribute interval numbers of environmental factors and different data-based models[J]. Catena, 2021, 202: 105250.

    [13] LIAO Lingxiao, LIU Jiamei, WANG Tao, et al. The application of the information model based on the symmetry method of disaster-causing factors in the risk assessment of earthquake-induced landslides [J]. Land and resources remote sensing, 2021, 33(2):172-181.(凌晓, 刘甲美, 王涛, 等. 基于致灾因子对称法分级的信息量模型在地震滑坡危险性评价中的应用[J]. 国土资源遥感, 2021, 33(2): 172-181.)
    [14] SUN Deliang, MA Xianglong, TANG Xiaoya, et al. Comparison of landslide susceptibility zoning based on different factor classifications - A case study of Wanzhou District [J]. Journal of Chongqing Normal University (Natural Science Edition), 2021, 38(05): 43-54.(孙德亮, 马祥龙, 唐小娅, 等. 基于不同因子分级的滑坡易发性区划对比—以万州区为例[J]. 重庆师范大学学报(自然科学版), 2021, 38(05): 43-54.)
    [15] SUN Deliang. Landslide susceptibility zoning and rainfall-induced landslide prediction and early warning based on machine learning [D]. Shanghai: East China Normal University, 2019.(孙德亮. 基于机器学习的滑坡易发性区划与降雨诱发滑坡预报预警研究[D]. 上海: 华东师范大学, 2019.)
    [16]

    Youssef A M, Pradhan B, Dikshit A, et al. Landslide susceptibility mapping using CNN-1D and 2D deep learning algorithms: Comparison of their performance at Asir Region, KSA[J]. Bulletin of Engineering Geology and the Environment, 2022, 81(4): 165.

    [17] ZHAO Zhan'ao, WANG Jizhou, MAO Xi, MA Weijun, LU Wenjuan, HE Yi, GAO Xuanyu. A multi-dimensional CNN coupled landslide susceptibility assessment method[J]. Geomatics and Information Science of Wuhan University, 2023. DOI: 10.13203/j.whugis20220325(赵占骜, 王继周, 毛曦, 马维军, 路文娟, 何毅, 高轩宇. 多维CNN耦合的滑坡易发性评价方法[J]. 武汉大学学报(信息科学版), 2023. DOI: 10.13203/j.whugis20220325)
    [18] CHEN Tao, ZHONG Ziying, NIU Ruiqing, LIU Tong, CHEN Shengyun. Mapping Landslide Susceptibility Based on Deep Belief Network[J]. Geomatics and Information Science of Wuhan University, 2020, 45(11): 1809-1817. DOI: 10.13203/j.whugis20190144(陈涛, 钟子颖, 牛瑞卿, 刘桐, 陈胜云. 利用深度信念网络进行滑坡易发性评价[J]. 武汉大学学报(信息科学版), 2020, 45(11): 1809-1817. DOI: 10.13203/j.whugis20190144)
    [19] GUO Fei, WANG Xiujuan, CHEN Xi, et al. Comparative analysis of susceptibility evaluation of small-scale landslide in Gannan area based on different models [J]. Chinese Journal of Geological Disasters and Prevention, 2022, 33(6): 125-133.(郭飞, 王秀娟, 陈玺, 等. 基于不同模型的赣南地区小型削方滑坡易发性评价对比分析[J]. 中国地质灾害与防治学报, 2022, 33(6): 125- 133.)
    [20] HUANG Faming, Li Jinfeng, WANG Junyu, et al. Modeling rules of landslide susceptibility prediction considering the suitability of linear environmental factors and different machine learning models [J]. Geological Science and Technology Bulletin, 2022, 41(2): 44-59.(黄发明, 李金凤, 王俊宇, 等. 考虑线状环境因子适宜性和不同机器学习模型的滑坡易发性预测建模规律[J]. 地质科技通报, 2022, 41(2): 44-59.)
    [21] ZHAO Jiubin, LIU Yuanxue, LIU Na, et al. Distributed BP neural network regional landslide spatial prediction method under massive monitoring data [J]. Geotechnical mechanics, 2019, 40(7): 2866-2872.(赵久彬, 刘元雪, 刘娜, 等. 海量监测数据下分布式BP神经网络区域滑坡空间预测方法[J]. 岩土力学, 2019, 40(7): 2866-2872.)
    [22] TIAN Naiman, LAN Hengxing, WU Yuming, et al. Performance comparison of artificial neural network and decision tree model in landslide susceptibility analysis [J]. Acta Geoinformatics, 2021, 22(12): 2304-2316.(田乃满, 兰恒星, 伍宇明, 等. 人工神经网络和决策树模型在滑坡易发性分析中的性能对比[J]. 地球信息科学学报, 2021, 22(12): 2304-2316.)
    [23] WU Hongyang, ZHOU Chao, LIANG Xin, et al. Landslide susceptibility assessment and zoning of Yanshan Township in the Three Gorges Reservoir Area based on XGBoost model[J]. Chinese Journal of Geological Hazard and Control, 2023, 34(05): 141- 152.(吴宏阳, 周超, 梁鑫, 等. 基于XGBoost模型的三峡库区燕山乡滑坡易发性评价与区划[J]. 中国地质灾害与防治学报, 2023, 34(05): 141-152.)
    [24]

    ZHOU X, WEN H, Li Z, et al. An interpretable model for the susceptibility of rainfall-induced shallow landslides based on SHAP and XGBoost[J]. Geocarto International, 2022, 37(26): 13419-13450.

    [25] LIU Yanhui, HUANG Junbao, XIAO Ruihua, et al. warning model of regional landslide disaster in Fujian Province based on random forest [J]. Journal of Engineering Geology, 2022, 30(3): 944-955.(刘艳辉, 黄俊宝, 肖锐铧, 等. 基于随机森林的福建省区域滑坡灾害预警模型研究[J]. 工程地质学报, 2022, 30(3): 944-955.)
    [26] ZHANG Wengang, HE Yuwei, WANG Luqi, et al. Machine learning analysis method of landslide susceptibility based on water system partition: taking Fengjie County of Chongqing as an example [J] Earth Science, 2023,48(5): 2024-2038. (仉文岗, 何昱苇, 王鲁琦, 等. 基于水系分区的滑坡易发性机器学习分析方法: 以重庆市奉节县为例[J]. 地球科学, 2023, 48(5): 2024-2038.)
    [27]

    KONG C, TIAN Y, MA X, et al. Landslide susceptibility assessment based on different machine learning methods in Zhaoping County of Eastern Guangxi[J]. Remote Sensing, 2021, 13(18): 3573.

    [28]

    JENNIFER J J. Feature elimination and comparison of machine learning algorithms in landslide susceptibility mapping[J]. Environmental Earth Sciences, 2022, 81(20): 1-23.

    [29] GUO Fei, JIANG Guanghui, HUANG Xiaohu, et al. Impact of environmental factor combinations and negative sample selection on Benggang susceptibility assessment in granite areas[J]. Transactions of the Chinese Society of Agricultural Engineering, 2024,40(1): 191-200. DOI: 10.11975/j.issn.1002-6819.202307270(郭飞, 蒋广辉, 黄晓虎, 等. 环境因子组合和负样本选取策略对花岗岩区崩岗易发性评价的影响[J]. 农业工程学报, 2024, 40(1): 191-200. DOI: 10.11975/j.issn.1002-6819.202307270)
    [30] GUO Fei, LAI Peng, HUANG Faming, et al. Literature review and research progress of landslide susceptibility evaluation based on knowledge map [J/OL]. Earth Science: 1-33[2024-2- 29].http://kns.cnki.net/kcms/detail/42.1874.P.20230713.1234.002.html.(郭飞,赖鹏,黄发明等.基于知识图谱的滑坡易发性评价文献综述及研究进展[J/OL].地球科学:1-33[2024-02-29].http://kns.cnki.net/kcms/detail/42.1874.P.20230713.1234.002.html.)
    [31] ZHAO Xiaoyan, TAN Shucheng, LI Yongping. Geological hazard risk assessment in Dongchuan District based on slope unit and combination weighting method [J]. Journal of Yunnan University (Natural Science Edition), 2021, 43(2): 299-305.(赵晓燕, 谈树成, 李永平. 基于斜坡单元与组合赋权法的东川区地质灾害危险性评价[J]. 云南大学学报(自然科学版), 2021, 43(2): 299- 305.)
    [32] WANG Jinfeng, XU Chengdong. Geodetector: Principles and Prospects [J]. Acta Geographica Sinica, 2017, 72(1).(, 徐成东. 地理探测器: 原理与展望[J]. 地理学报, 2017, 72(1).)
    [33]

    RUMELHART D E, HINTON G E, WILLIAMS R J. Learning representations by back-propagating errors[J]. Nature, 1986, 323(6088): 533-536.

    [34]

    CHEN T, GUESTRIN C. Xgboost: A scalable tree boosting system[C]//Proceedings of the 22nd acm sigkdd international conference on knowledge discovery and data mining. 2016: 785-794.

    [35]

    BREIMAN L. Random forests[J]. Machine learning, 2001, 45: 5-32.

    [36] CHEN Yuping, YUAN Zhiqiang, ZHOU Bo, et al. Application of BP network optimized by genetic algorithm in landslide disaster prediction [J]. Hydrogeological Engineering Geology, 2012,39(1): 114-119.(陈玉萍, 袁志强, 周博, 等. 遗传算法优化BP网络在滑坡灾害预测中的应用研究[J]. 水文地质工程地质, 2012, 39(1): 114-119.)
    [37] LIU Jian, LI Shulin, CHEN Tao. Landslide susceptibility assessment based on optimized random forest model [J]. Journal of Wuhan University · Information Science Edition, 2018, 43(7): 1085-1091.(刘坚, 李树林, 陈涛. 基于优化随机森林模型的滑坡易发性评价[J]. 武汉大学学报(信息科学版), 2018, 43(7):1085-1091.)
    [38] HONG Zenglin, LI Yonghong, ZHANG Lingyu, et al. A regional geological hazard risk assessment method based on principal component analysis [J]. Disaster Science, 2020,35(1): 118-124.12.(洪增林, 李永红, 张玲玉, 等. 一种基于主成分分析法的区域性地质灾害危险性评估方法[J]. 灾害学, 2020, 35(1): 118-124.12.)
    [39] Meng Fanqi, Gao Feng, Lin Bo, et al. Geological hazard susceptibility evaluation based on AHP and information quantity models: A case study of the Ludong area [J]. Disaster Science, 2023,38(03): 111-117.(孟凡奇, 高峰, 林波, 等. 基于AHP和信息量模型的地质灾害易发性评价——以鲁东片区为例[J]. 灾害学, 2023, 38(3): 111-117.)
    [40] YI Sicai, ZHANG Mingwen. Development characteristics and risk assessment of debris flow in a village of Banpo Township, Luchun County [J].Geological Disasters and Environmental Protection, 2022,33(1): 15-23.(易思材,张明文.绿春县半坡乡某村泥石流发育特征及危险性评价[J].地质灾害与环境保护, 2022, 33(1): 15-23.)
    [41] Huang Zhijie, Jian Wenbin, Xia Chang et al. A prediction method for displacement rate of step type landslides based on LSO-RF model [J]. Journal of Fuzhou University (Natural Science Edition), 2023,51(06): 872-878(黄智杰, 简文彬, 夏昌, 等. 基于LSORF模型的阶跃型滑坡位移速率预测方法[J]. 福州大学学报(自然科学版), 2023, 51(6): 872-878.)
    [42] Ministry of Natural Resources of the People's Republic of China. DZ/T 0438-2023, Standard for Geohazard Risk Survey and Evaluation (1:50000) [S]. Beijing: Ministry of Natural Resources, 2023.(中华人民共和国自然资源部. DZ/T 0438-2023, 地质灾害风险调查评价规范(1:50000)[S]. 北京: 自然资源部, 2023.)
    [43] HUANG Faming, PAN Lihan, YAO Chi, et al. Landslide susceptibility prediction modeling based on semi-supervised machine learning [J]. Journal of Zhejiang University (Engineering Edition), 2021, 55(09): 1705-1713.(黄发明, 潘李含, 姚池, 等. 基于半监督机器学习的滑坡易发性预测建模[J]. 浙江大学学报(工学版), 2021, 55(09): 1705-1713.)
计量
  • 文章访问数:  90
  • HTML全文浏览量:  14
  • PDF下载量:  19
  • 被引次数: 0
出版历程
  • 收稿日期:  2024-09-10

目录

    /

    返回文章
    返回