基于集成学习建模的滑坡易发性评价

邬礼扬; 曾韬睿; 刘谢攀; 郭子正; 刘真意; 殷坤龙

doi:10.3799/dqkx.2022.451

基于集成学习建模的滑坡易发性评价

doi: 10.3799/dqkx.2022.451

邬礼扬^1, ,,
曾韬睿³,
刘谢攀²,
郭子正⁴,
刘真意⁵,
殷坤龙^2, , ,

1.
湖北省地质灾害防治中心, 湖北武汉 430030
2.
中国地质大学工程学院, 湖北武汉 430074
3.
中国地质大学地质调查研究院, 湖北武汉 430074
4.
河北工业大学土木与交通学院, 天津 300401
5.
中铁二院工程集团有限责任公司, 四川成都 610031

基金项目:

国家重点研发计划项目 2018YFC0809402

国家自然科学基金项目 41877525

详细信息

作者简介:
邬礼扬（1998-），男，硕士研究生，主要从事地质灾害预测预报和风险分析方面的研究. ORCID：0000-0003-0181-9168. E-mail：wuliyang@cug.edu.cn

通讯作者:
殷坤龙(1963-)，男，教授, ORCID: 0000-0002-3547-1633. E-mail: yinkl@126.com

中图分类号: P642.22
计量
- 文章访问数: 501
- HTML全文浏览量: 153
- PDF下载量: 83
- 被引次数: 0
出版历程
- 收稿日期: 2022-07-31
- 网络出版日期: 2024-11-08
- 刊出日期: 2024-10-25

Landslide Susceptibility Assessment Based on Ensemble Learning Modeling

Wu Liyang^{1
,
,},
Zeng Taorui³,
Liu Xiepan²,
Guo Zizheng⁴,
Liu Zhenyi⁵,
Yin Kunlong^{2
, ,
,}

1.
Hubei Center of Geological Disaster Control, Wuhan 430030, China
2.
Faculty of Engineering, China University of Geosciences, Wuhan 430074, China
3.
Institute of Geological Survey, China University of Geosciences, Wuhan 430074, China
4.
School of Civil and Transportation Engineering, Hebei University of Technology, Tianjin 300401, China
5.
China Railway Eryuan Engineering Group Co., Ltd., Chengdu 610031, China

摘要

摘要: 单一的机器学习模型往往难以满足滑坡易发性制图的需要，为提升滑坡易发性评价精度.提出了一种基于集成策略的机器学习模型组合优化的方法，以重庆市云阳县西部的12个乡镇为例进行滑坡易发性评价.首先，基于366处滑坡数据以及高程、坡度等9个指标因子构建易发性评价指标体系.然后以决策树模型（decision tree mode，DT）、逻辑回归模型（logistic regression，LR）和贝叶斯网络模型（bayesian network，BN）为基础模型，利用集成学习的三大类型，袋装法（bagging）、提升法（boosting）以及堆叠法（stacking）进行模型组合.并对各组合模型分别用粒子群算法（particle swarm optimization，PSO），贝叶斯算法（bayesian optimization，BO）进行超参数优化以及K最邻近算法（K-nearest neighbor，KNN）进行模型链接.最后采用ROC曲线与统计分析的方式来评估各集成学习模型精度.研究结果表明：与基础模型相比，三类集成学习模型精度均有提升，DT-LR-BN模型提升了3.5%~12.8%，RF模型提升了8%；BO-XGBoost模型提升了13.1%；KNN-stacking模型提升了7.4%~17%，BO-XGBoost模型的AUC值最高为0.811.集成学习能有效提升机器学习模型性能，提高滑坡易发性制图的精度，研究为机器学习模型之间的组合优化提供了新的思路与方法.
- 集成学习 /
- 袋装法 /
- 提升法 /
- 堆叠法 /
- 滑坡灾害 /
- 易发性评价 /
- 工程地质学
Abstract: A single machine learning model is often difficult to meet the needs of landslide vulnerability mapping, in order to improve the accuracy of landslide vulnerability assessment. In this paper, a method of machine learning model combination optimization based on integrated strategy is proposed, twelve townships in the west of Yunyang County, Chongqing were taken as an example. First, based on 366 landslide data and 9 index factors such as elevation and slope, the susceptibility evaluation index system was constructed. Then used the three algorithms of ensemble learning, bagging, boosting and stacking, to build combined models based on Decision Tree Mode (DT), Logic Regression (LR) and Bayesian Network (BN).The combined models used Particle Swarm Optimization (PSO), Bayesian Optimization (BO) for super parameter optimization and K-Nearest Neighbor (KNN) was used for model recombination. Finally, ROC curve and statistical analysis were used to calculate the accuracy of each integrated learning model. The research results show that compared with the basic classifier models, the accuracy of the three types of integrated learning models was improved. the DT-LR-BN model increased by 3.5%-12.8%, the RF model increased by 8%; the BO-XGBoost model increased by 13.1%; the KNN-stacking model increased by 7.4%-17%, and the AUC value of BO-XGBoost model was the highest at 0.811. Integrated learning can effectively improve the performance of machine learning models, improve the accuracy of landslide susceptibility mapping, and provide a new idea and method for the combination optimization between machine learning models.
- ensemble learning /
- bagging /
- boosting /
- stacking /
- landslide disaster /
- susceptibility evaluation /
- engineering geology

HTML全文

图 1 研究区地理位置(a), 滑坡灾害分布(b), 蔡家坝滑坡范围(c)

Fig. 1. Geographical location of the study area (a), landslide hazard distribution (b), Caijiaba landslide range (c)

下载: 全尺寸图片幻灯片

图 2 指标因子相关性矩阵

Fig. 2. Index factor correlation matrix

下载: 全尺寸图片幻灯片

图 3 指标因子分区

Fig. 3. Index factor partition diagrams

下载: 全尺寸图片幻灯片

图 4 集成学习建模技术路线图

a. Bagging模型；b. BO-Boosting模型；c. KNN-Stacking模型

Fig. 4. Road map of ensemble learning modeling technology

下载: 全尺寸图片幻灯片

图 5 PSO算法迭代过程

Fig. 5. Iterative process of PSO algorithm

下载: 全尺寸图片幻灯片

图 6 集成学习模型ROC曲线

a. Bagging模型；b. BO-XGBoost模型；c. KNN-stacking模型

Fig. 6. ROC curves of ensemble learning model

下载: 全尺寸图片幻灯片

图 7 蔡家坝滑坡统计

a. 基础模型；b. 集成学习模型

Fig. 7. Statistical chart of Caijiaba landslide

下载: 全尺寸图片幻灯片

图 8 滑坡易发性区划

a. DT模型；b. LR模型；c. BN模型；d. DT-LR-BN模型；e. RF模型；f. KNN-Stacking模型；g. BO-XGBoost模型

Fig. 8. Landslide susceptibility zoning map

下载: 全尺寸图片幻灯片

表 1 BO-XGBoost模型参数

Table 1. BO-XGBoost model parameters

参数名称	参数空间	最终取值
决策树数量	[100, 300]	125
学习率	[0.01, 1]	0.05
决策树最大深度	[4, 20]	12
每次生成树时随机抽样因子比例	[0, 1]	1
每次生成树时随机抽样特征比例	[0, 1]	0.7
每次生成叶子节点时随机抽样特征比例	[0, 1]	1
随机抽取样本比例	[0, 1]	0.9

下载: 导出CSV

表 2 模型栅格统计

Table 2. Model grid statistics

模型	各易发性等级总栅格数				各易发性等级滑坡栅格数				滑坡比率
模型	低	中	高	极高	低	中	高	极高	低	中	高	极高
DT	420 157	188 796	117 311	167 708	4 045	5 942	7 996	17 891	0.24	0.78	1.70	2.66
LR	165 465	184 256	194 882	349 369	730	2 610	5 313	27 221	0.11	0.35	0.68	1.94
BN	214 077	257 605	253 802	168 488	2 916	7 733	13 093	12 132	0.34	0.75	1.29	1.79
RF	291 291	222 377	207 536	172 768	1 414	4 217	8 934	21 309	0.12	0.47	1.07	3.07
DT-LR-BN	217 441	279 704	227 288	169 539	1 094	4 762	10 433	19 585	0.13	0.42	1.14	2.88
BO-XGBoost	243 189	232 152	234 475	184 156	473	3 329	8 860	23 212	0.05	0.36	0.94	3.14
KNN-stacking	397 928	170 540	150 620	174 884	4 033	5 024	7 428	19 389	0.25	0.73	1.23	2.76

下载: 导出CSV

参考文献(63)

Ankita, Sahana, S. K., 2021. Ba-PSO: A Balanced PSO to Solve Multi-Objective Grid Scheduling Problem. Applied Intelligence, 52(4): 4015-4027. https://doi.org/10.1007/s10489-021-02625-7
Ayalew, L., Yamagishi, H., 2005. The Application of GIS-Based Logistic Regression for Landslide Susceptibility Mapping in the Kakuda-Yahiko Mountains, Central Japan. Geomorphology, 65(1-2): 15-31. https://doi.org/10.1016/j.geomorph.2004.06.010
Can, R., Kocaman, S., Gokceoglu, C., 2021. A Comprehensive Assessment of XGBoost Algorithm for Landslide Susceptibility Mapping in the Upper Basin of Ataturk Dam, Turkey. Applied Sciences, 11(11): 4993. https://doi.org/10.3390/app11114993
Du, C. J., Yi, Q. L., Zhou, B., et al., 2017. Landslide Susceptibility Assessment in Yunyang District of Three Gorges Reservoir Based on GIS and Weighted Information Value. Journal of China Three Gorges University (Natural Sciences), 39(2): 48-53 (in Chinese with English abstract).
Du, G. L., Yang, Z. H., Yuan, Y., et al., 2021. Landslide Susceptibility Mapping in the Sichuan-Tibet Traffic Corridor Using Logistic Regression-Information Value Method. Hydrogeology & Engineering Geology, 48(5): 102-111 (in Chinese with English abstract).
Fang, Z. C., Wang, Y., Peng, L., et al., 2021. A Comparative Study of Heterogeneous Ensemble-Learning Techniques for Landslide Susceptibility Mapping. International Journal of Geographical Information Science, 35(2): 321-347. https://doi.org/10.1080/13658816.2020.1808897
Guo, Z. Z., Yin, K. L., Huang, F. M., et al., 2019. Landslide Susceptibility Evaluation Based on Landslide Classification and Weighted Frequency Ratio Model. Chinese Journal of Rock Mechanics and Engineering, 38(2): 287-300 (in Chinese with English abstract).
Han, J. C., Zhang, H., Cao, J., 2021. Assessing Earthquake-Induced Landslide Susceptibility Based on Logistic Regression in 2008 Wenchuan Earthquake and 2014 Ludian Earthquake. Journal of Catastrophology, 36(2): 193-199 (in Chinese with English abstract). doi: 10.3969/j.issn.1000-811X.2021.02.034
Huang, F. M., Yin, K. L., Jiang, S. H., et al., 2018. Landslide Susceptibility Assessment Based on Clustering Analysis and Support Vector Machine. Chinese Journal of Rock Mechanics and Engineering, 37(1): 156-167 (in Chinese with English abstract).
Huang, W. B., Ding, M. T., Wang, D., et al., 2022. Evaluation of Landslide Susceptibility Based on Layer Adaptive Weighted Convolutional Neural Network Model along Sichuan-Tibet Traffic Corridor. Earth Science, 47(6): 2015-2030 (in Chinese with English abstract).
Jurek, A., Bi, Y. X., Wu, S. L., et al., 2014. A Survey of Commonly Used Ensemble-Based Classification Techniques. The Knowledge Engineering Review, 29(5): 551-581. https://doi.org/10.1017/s0269888913000155
Kavzoglu, T., Teke, A., 2022. Advanced Hyperparameter Optimization for Improved Spatial Prediction of Shallow Landslides Using Extreme Gradient Boosting (XGBoost). Bulletin of Engineering Geology and the Environment, 81(5): 1-22. https://doi.org/10.1007/s10064-022-02708-w
Li, J. J., Wang, H., Wang, R., 2022. Landslide Susceptibility Assessment Based on Machine Learning in Fuzhou Area. Journal of Water Resources and Architectural Engineering, 20(2): 146-152 (in Chinese with English abstract).
Li, W. B., Fan, X. M., Huang, F. M., et al., 2021. Uncertainties of Landslide Susceptibility Modeling under Different Environmental Factor Connections and Prediction Models. Earth Science, 46(10): 3777-3795 (in Chinese with English abstract).
Liu, J., Li, S. L., Chen, T., 2018. Landslide Susceptibility Assesment Based on Optimized Random Forest Model. Geomatics and Information Science of Wuhan University, 43(7): 1085-1091 (in Chinese with English abstract).
Liu, S. H., Yin, K. L., Zhou, C., et al., 2021. Susceptibility Assessment for Landslide Initiated along Power Transmission Lines. Remote Sensing, 13(24): 5068. https://doi.org/10.3390/rs13245068
Mancini, F., Ceppi, C., Ritrovato, G., 2010. GIS and Statistical Analysis for Landslide Susceptibility Mapping in the Daunia Area, Italy. Natural Hazards and Earth System Sciences, 10(9): 1851-1864. https://doi.org/10.5194/nhess-10-1851-2010
Sheng, M. Q., Liu, Z. X., Zhang, X. Q., et al., 2021. Landslide Susceptibility Prediction Based on Frequency Ratio Analysis and Support Vector Machine. Science Technology and Engineering, 21(25): 10620-10628 (in Chinese with English abstract). doi: 10.3969/j.issn.1671-1815.2021.25.009
Song, Y, Gong, J, Gao, S, et al., 2012. Susceptibility Assessment of Earthquake-Induced Landslides Using Bayesian Network: A Case Study in Beichuan, China. Computers & Geosciences, 42: 189-199. https://doi.org/10.1016/j.cageo.2011.09.011
Tehrany, M. S., Pradhan, B., Jebur, M. N., 2013. Spatial Prediction of Flood Susceptible Areas Using Rule Based Decision Tree (DT) and a Novel Ensemble Bivariate and Multivariate Statistical Models in GIS. Journal of Hydrology, 504: 69-79. https://doi.org/10.1016/j.jhydrol.2013.09.034
Tian, N. M., Lan, H. X., Wu, Y. M., et al., 2020. Performance Comparison of BP Artificial Neural Network and CART Decision Tree Model in Landslide Susceptibility Prediction. Journal of Geo-Information Science, 22(12): 2304-2316 (in Chinese with English abstract).
Wang, J. J., Yin, K. L., Xiao, L. L., 2014. Landslide Susceptibility Assessment Based on GIS and Weighted Information Value: A Case Study of Wanzhou District, Three Gorges Reservoir. Chinese Journal of Rock Mechanics and Engineering, 33(4): 797-808 (in Chinese with English abstract).
Wang, Q., Xue, Y., Zhang, W., et al., 2021. Landslide Susceptibility Mapping Based on Support Vector Machine Models. Journal of Hunan City University (Natural Science), 30(1): 22-28 (in Chinese with English abstract).
Wang, S. B., Zhuang, J. Q., Zheng, J., et al., 2021. Application of Bayesian Hyperparameter Optimized Random Forest and XGBoost Model for Landslide Susceptibility Mapping. Frontiers in Earth Science, 9: 712240. https://doi.org/10.3389/feart.2021.712240
Wei, A. H., Yu, K. N., Dai, F. G., et al., 2022. Application of Tree-Based Ensemble Models to Landslide Susceptibility Mapping: A Comparative Study. Sustainability, 14(10): 6330. https://doi.org/10.3390/su14106330
Xie, W. Y., Chai, Q. Q., Gan, Y. H., et al., 2020. Strains Classification of Anoectochilus Roxburghii Using Multi-Feature Extraction and Stacking Ensemble Learning. Transactions of the Chinese Society of Agricultural Engineering, 36(14): 203-210 (in Chinese with English abstract).
Xu, J. W., Yang, Y., 2018. A Survey of Ensemble Learning Approaches. Journal of Yunnan University (Natural Sciences Edition), 40(6): 1082-1092 (in Chinese with English abstract).
Xu, S. H., Liu, J. P., Wang, X. H., et al., 2020. Landslide Susceptibility Assessment Method Incorporating Index of Entropy Based on Support Vector Machine: A Case Study of Shaanxi Province. Geomatics and Information Science of Wuhan University, 45(8): 1214-1222 (in Chinese with English abstract).
Yang, S., Li, D. Y., Yan, L. X., et al., 2021. Landslide Susceptibility Assessment in High and Steep Bank Slopes along Wujiang River Based on Random Forest Model. Safety and Environmental Engineering, 28(4): 131-138 (in Chinese with English abstract).
Yang, Y. G., Yin, K. L., Zhao, H. Y., et al., 2019. Landslide Susceptibility Evaluation for Township Units of Bank Section in Wanzhou District Based on C5.0 Decision Tree and K-Means Cluster Model. Geological Science and Technology Information, 38(6): 189-197 (in Chinese with English abstract).
Yin, K. L., Zhu, L. F., 2001. Landslide Hazard Zonation and Application of GIS. Earth Science Frontiers, 8(2): 279-284 (in Chinese with English abstract).
Zeng, T. R., Jiang, H. W., Liu, Q. L., et al., 2022. Landslide Displacement Prediction Based on Variational Mode Decomposition and MIC-GWO-LSTM Model. Stochastic Environmental Research and Risk Assessment, 36(5): 1353-1372. https://doi.org/10.1007/s00477-021-02145-3
Zhang, J., Yin, K. L., Wang, J. J., et al., 2016. Evaluation of Landslide Susceptibility for Wanzhou District of Three Gorges Reservoir. Chinese Journal of Rock Mechanics and Engineering, 35(2): 284-296 (in Chinese with English abstract).
Zhang, S. C., Li, X. L., Zong, M., et al., 2017. Learning k for kNN Classification. ACM Transactions on Intelligent Systems and Technology, 8(3): 1-19. https://doi.org/10.1145/2990508
Zhang, Y. Y., Wen, H. J., Ma, C. C., et al., 2018. Study on Genetic Mechanism and Stability Evaluation of Caijiaba Super-Large Landslide Based on Multi-Source Data. Chinese Journal of Rock Mechanics and Engineering, 37(9): 2048-2063 (in Chinese with English abstract).
Zhao, D. M., Xie, J. K., Wang, C., et al., 2022. On-Line Transient Stability Assessment of a Power System Based on Bagging Ensemble Learning. Power System Protection and Control, 50(8): 1-10 (in Chinese with English abstract).
Zhou, C., Yin, K. L., Cao, Y., et al., 2020. Landslide Susceptibility Assessment by Applying the Coupling Method of Radial Basis Neural Network and Adaboost: A Case Study from the Three Gorges Reservoir Area. Earth Science, 45(6): 1865-1876 (in Chinese with English abstract).
Zhou, W. H., Zhai, X. F., Tan, H. W., 2022. Research on Financial Frauds Prediction Model of Chinese Public Companies with XGBoost. The Journal of Quantitative & Technical Economics, 39(7): 176-196 (in Chinese with English abstract).
Zounemat-Kermani, M., Batelaan, O., Fadaee, M., et al., 2021. Ensemble Machine Learning Paradigms in Hydrology: A Review. Journal of Hydrology, 598: 126266. https://doi.org/10.1016/j.jhydrol.2021.126266
杜常见, 易庆林, 周宝, 等, 2017. 基于GIS和加权信息量的三峡库区云阳县滑坡灾害易发性评价. 三峡大学学报(自然科学版), 39(2): 48-53.
杜国梁, 杨志华, 袁颖, 等, 2021. 基于逻辑回归-信息量的川藏交通廊道滑坡易发性评价. 水文地质工程地质, 48(5): 102-111.
郭子正, 殷坤龙, 黄发明, 等, 2019. 基于滑坡分类和加权频率比模型的滑坡易发性评价. 岩石力学与工程学报, 38(2): 287-300.
韩继冲, 张朝, 曹娟, 2021. 基于逻辑回归的地震滑坡易发性评价: 以汶川地震、鲁甸地震为例. 灾害学, 36(2): 193-199.
黄发明, 殷坤龙, 蒋水华, 等, 2018. 基于聚类分析和支持向量机的滑坡易发性评价. 岩石力学与工程学报, 37(1): 156-167.
黄武彪, 丁明涛, 王栋, 等, 2022. 基于层数自适应加权卷积神经网络的川藏交通廊道沿线滑坡易发性评价. 地球科学, 47(6): 2015-2030.
李嘉靖, 王浩, 王睿, 2022. 基于机器学习模型的福州地区滑坡易发性评价. 水利与建筑工程学报, 20(2): 146-152.
李文彬, 范宣梅, 黄发明, 等, 2021. 不同环境因子联接和预测模型的滑坡易发性建模不确定性. 地球科学, 46(10): 3777-3795. doi: 10.3799/dqkx.2021.042
刘坚, 李树林, 陈涛, 2018. 基于优化随机森林模型的滑坡易发性评价. 武汉大学学报(信息科学版), 43(7): 1085-1091.
盛明强, 刘梓轩, 张晓晴, 等, 2021. 基于频率比联接法和支持向量机的滑坡易发性预测. 科学技术与工程, 21(25): 10620-10628.
田乃满, 兰恒星, 伍宇明, 等, 2020. 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. 地球信息科学学报, 22(12): 2304-2316.
王佳佳, 殷坤龙, 肖莉丽, 2014. 基于GIS和信息量的滑坡灾害易发性评价: 以三峡库区万州区为例. 岩石力学与工程学报, 33(4): 797-808.
王倩, 薛云, 张维, 等, 2021. 基于支持向量机的滑坡易发性评价. 湖南城市学院学报(自然科学版), 30(1): 22–28.
谢文涌, 柴琴琴, 甘勇辉, 等, 2020. 基于多特征提取和Stacking集成学习的金线莲品系分类. 农业工程学报, 36(14): 203-210.
徐继伟, 杨云, 2018. 集成学习方法: 研究综述. 云南大学学报(自然科学版), 40(6): 1082-1092.
徐胜华, 刘纪平, 王想红, 等, 2020. 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. 武汉大学学报(信息科学版), 45(8): 1214-1222.
杨硕, 李德营, 严亮轩, 等, 2021. 基于随机森林模型的乌江高陡岸坡滑坡地质灾害易发性评价. 安全与环境工程, 28(4): 131-138.
杨永刚, 殷坤龙, 赵海燕, 等, 2019. 基于C5.0决策树-快速聚类模型的万州区库岸段乡镇滑坡易发性区划. 地质科技情报, 38(6): 189-197.
殷坤龙, 朱良峰, 2001. 滑坡灾害空间区划及GIS应用研究. 地学前缘, 8(2): 279-284.
张俊, 殷坤龙, 王佳佳, 等, 2016. 三峡库区万州区滑坡灾害易发性评价研究. 岩石力学与工程学报, 35(2): 284-296.
张岩岩, 文海家, 麻超超, 等, 2018. 基于多源数据的蔡家坝特大型滑坡成因机制研究及稳定性评价. 岩石力学与工程学报, 37(9): 2048-2063.
赵冬梅, 谢家康, 王闯, 等, 2022. 基于Bagging集成学习的电力系统暂态稳定在线评估. 电力系统保护与控制, 50(8): 1-10.
周超, 殷坤龙, 曹颖, 等, 2020. 基于集成学习与径向基神经网络耦合模型的三峡库区滑坡易发性评价. 地球科学, 45(6): 1865-1876. doi: 10.3799/dqkx.2020.071
周卫华, 翟晓风, 谭皓威, 2022. 基于XGBoost的上市公司财务舞弊预测模型研究. 数量经济技术经济研究, 39(7): 176-196.