基于优化负样本采样策略的梯度提升决策树与随机森林的汶川同震滑坡易发性评价

摘要:
强震诱发的滑坡具有数量多、分布广、规模大等特点, 严重威胁人民生命财产安全。滑坡易发性评价能够快速预测灾害空间分布, 对于减轻震后灾害的危险性具有重要意义。在同震滑坡易发性评价研究中, 如何选取滑坡负样本并通过耦合机器学习模型提高评价精度的对比研究仍需进一步研究。以山区汶川地震诱发的滑坡为研究区, 首先选取地形地貌、地质环境、地震参数等10个滑坡评价因子, 分析滑坡空间分布规律; 其次因子共线性分析检验数据冗余, 接下来采用频率比法(FR)选取极低、低易发区滑坡负样本点的采样策略; 最后采用基于决策树演化改进的梯度提升决策树(GBDT)、随机森林(RF)和耦合模型(FR-GBD与FR-RF), 开展了基于机器学习的同震滑坡易发性区划并进行精度评价。研究结果表明: ①滑坡空间分布受到多层级因子控制; ②模型预测精度为: FR-RF(AUC=0.943) >FR-GBDT(AUC=0.926)>RF(AUC=0.901)>GBDT(AUC=0.856);③在低易发区选择滑坡负样本可以明显提高易发性精度。研究成果可为滑坡易发性中负样本的选择和评价模型构建提供参考同时也为震后滑坡的防灾减灾提供理论支持。
- 随机森林(RF) /
- 梯度提升决策树(GBDT) /
- 机器学习 /
- 频率比法(FR) /
- 采样策略 /
- 同震滑坡 /
- 滑坡易发性区划
Abstract: Objective
Strong earthquake-induced landslides are characterized by large number, wide distribution and large scale, and seriously threaten people's lives and property. Landslide susceptibility mapping (LSM) can quickly predict the spatial distribution of prone areas, which is highly important for reducing the risk of post-earthquake disasters. However, in the studies of coseismic landslide LSMs, how to select negative landslide samples and integrate machine learning models to improve the evaluation accuracy still needs further investigation.
Methods
In this study, the landslides induced by the Wenchuan earthquake in mountainous areas are selected as a case study. First, 10 landslide influencing factors, such as topography, geological environment, and seismic parameters, are selected to analyse the spatial distribution of landslides. Then, collinearity analysis is used to test data redundancy, nonnegative sample points from the sampling strategies are randomly selected in the extremely low susceptibility regions by the frequency ratio (FR) method. Finally, gradient boosting decision tree (GBDT), random forest (RF), and their optimal models are used to predict coseismic landslide susceptibility, conduct a comparative study of the models and carry out an accuracy assessment.
Results
The results show that ① the spatial distribution of landslides is controlled by multiple factors, and ② the accuracy of the models is FR-RF(AUC=0.943)>FR-GBDT(AUC=0.926)>RF(AUC=0.901)>GBDT(AUC=0.856). ③ Selecting negative landslide samples in low susceptibility areas could significantly improve the accuracy of LSMs.
Conclusion
The research results can provide a reference for selecting negative landslide samples and constructing evaluation models, as well as for providing theoretical support for post-earthquake disaster prevention and mitigation.
- random forest(RF) /
- gradient boosting decision tree(GBDT) /
- machine learning /
- frequency ratio(FR) /
- sampling strategy /
- coseismic landslide /
- landslide susceptibility mapping
注释:

所有作者声明不存在利益冲突。

HTML全文

图 1 研究区概况及滑坡分布(图b、c为局部放大图)

Figure 1. General situation of the study area and spatial distribution of landslids

下载: 全尺寸图片幻灯片

图 2 滑坡致灾评价分级图

PGA.地面峰值加速度；PGV.地面峰值速度，下同

Figure 2. Classification of landslide evaluation factors

下载: 全尺寸图片幻灯片

图 3 技术路线

Figure 3. Technological flow chart

下载: 全尺寸图片幻灯片

图 4 基于决策树的机器学习模型

Figure 4. Machine learning models based on decision tree

下载: 全尺寸图片幻灯片

图 5 汶川地震滑坡发生频率与评价因子的相关性

Figure 5. Correlation between Wenchuan earthquake-induced landslide frequency and evaluation factors

下载: 全尺寸图片幻灯片

图 6 频率比模型易发性

Figure 6. Landslide susceptibility map of frequency ratio model

下载: 全尺寸图片幻灯片

图 7 优化机器学习模型易发性图

Figure 7. Landslide susceptibility map of optimized machine learning models

下载: 全尺寸图片幻灯片

图 8 机器学习模型易发性分级

Figure 8. Classification of landslide susceptibility for machine learning models

下载: 全尺寸图片幻灯片

图 9 ROC曲线和AUC值

Figure 9. Receiver operating characteristic curves and area under curve values

下载: 全尺寸图片幻灯片

图 10 基于最优模型的评价指标特征重要性

Figure 10. Feature importance of evaluation indicators based on optimal model

下载: 全尺寸图片幻灯片

表 1 评价因子数据来源

Table 1. Evaluation factors data source

因子类别	数据来源	分辨率
高程(DEM)	https://search.asf.alaska.edu	30 m
曲率	DEM
坡向	DEM
坡度	DEM
距水系距离	DEM
地层岩性	地质图	1∶200 000
距断层距离	地质图	1∶200 000
距道路距离	91卫图-全国矢量路网
PGA	https://www.usgs.gov/programs/earthquake-hazards
PGV	https://www.usgs.gov/programs/earthquake-hazards

下载: 导出CSV

表 2 研究区滑坡地质灾害评价因子指标分级

Table 2. Index classification of evaluation factors of landslides in study area

评价因子	指标分级	类别面积/km²	滑坡面积/km²	频率比
高程/km	[0, 1.0)	49.768 2	2.398 5	0.352 2
	[1.0, 1.5)	178.037 1	34.538 4	1.417 5
	[1.5, 2.0)	213.471 0	44.883 9	1.536 4
	[2.0, 2.5)	238.621 5	35.127 9	1.075 7
	[2.5, 3.0)	191.203 2	16.661 7	0.636 7
	[3.0, 3.5]	107.064 9	4.770 0	0.325 5
	＞3.5	37.253 7	0.584 1	0.114 6
坡度/(°)	[0, 10)	49.443 3	3.577 5	0.528 7
	[10, 20)	121.149 9	9.950 4	0.600 1
	[20, 30)	248.157 0	23.891 4	0.703 5
	[30, 40)	353.141 1	45.912 6	0.950 0
	[40, 50]	227.951 1	41.776 2	1.339 1
	＞50	64.165 5	16.715 7	1.903 6
坡向	平地	0.025 2	0.000 0	0.000 0
	北	107.894 7	12.094 2	0.819 1
	东北	112.387 5	13.443 3	0.874 0
	东	135.488 7	18.674 1	1.007 1
	东南	158.031 9	25.788 6	1.192 4
	南	124.151 4	22.896 0	1.347 6
	西南	113.337 9	18.488 7	1.192 0
	西	126.403 2	14.545 8	0.840 9
	西北	137.699 1	13.180 5	0.699 4
曲率	[0, 0.48)	186.871 5	22.210 2	0.868 5
	[0.48, 1.24)	492.058 8	62.758 8	0.932 0
	[1.24, 2.23)	311.374 8	45.416 7	1.065 8
	[2.23, 3.57)	81.992 7	15.079 5	1.343 9
	[3.57, 8.75]	14.014 8	3.131 1	1.632 5
地层岩性	千枚岩	112.330 8	27.315 9	1.776 9
	变质碳酸盐岩	35.205 3	3.835 8	0.796 1
	含砾细砂岩	1.596 6	0.000 0	0.000 0
	斜长角闪岩	211.251 6	45.921 6	1.588 4
	早元古代火成岩	7.509 6	1.903 5	1.852 2
	泥沙质岩	1.701 0	0.000 0	0.000 0
	火山岩	114.417 9	3.163 5	0.202 0
	灰岩	31.532 4	4.103 1	0.950 8
	砂质黏土	2.323 8	0.000 0	0.000 0
	砾岩	4.502 7	0.000 0	0.000 0
	碳酸盐岩	1.798 2	0.000 0	0.000 0
	花岗岩	391.212 9	43.580 7	0.814 0
	辉长岩	10.849 5	4.170 6	2.808 9
	酸性凝灰岩	19.157 4	2.597 4	0.990 7
	长石砂岩	103.202 1	2.618 1	0.185 4
距断层距离/km	[0, 4)	300.789 0	72.466 2	1.760 4
	[4, 8)	137.484 9	19.540 8	1.038 6
	[8, 12)	95.306 4	6.501 6	0.498 5
	[12, 17)	93.276 9	6.497 1	0.509 0
	[17, 22]	92.567 7	6.011 1	0.474 5
	＞22	129.155 4	4.027 5	0.227 9
距水系距离/km	[0, 2)	258.900 3	49.788 9	1.405 2
	[2, 4)	235.117 8	37.761 3	1.173 6
	[4, 6)	174.542 4	21.051 9	0.881 3
	[6, 8)	105.759 0	5.688 0	0.393 0
	[8, 10]	57.138 3	0.754 2	0.096 4
	＞10	47.133 9	0.000 0	0.000 0
距道路距离/km	[0, 1)	187.023 6	28.274 4	1.104 7
	[1, 2)	160.769 7	24.274 8	1.103 3
	[2, 4)	243.599 4	34.794 9	1.043 7
	[4, 6)	151.361 1	20.677 5	0.998 2
	[6, 8)	76.136 4	6.399 0	0.614 1
	[8, 10]	27.782 1	0.623 7	0.164 0
	＞10	1.908 0	0.000 0	0.000 0
地面峰值速度PGV/(cm·s^-1)	[0, 20)	87.277 5	2.106 0	0.176 3
	[20, 26)	67.586 4	3.358 8	0.363 1
	[26, 32)	151.925 4	2.471 4	0.118 9
	[32, 38)	239.892 3	21.619 8	0.658 5
	[38, 44]	313.504 2	57.429 9	1.338 6
	＞44	228.259 8	37.093 5	1.187 4
地面峰值加速度PGA/g	[0, 0.46)	87.277 5	5.594 4	0.468 4
	[0.46, 0.56)	67.586 4	21.789 9	2.355 8
	[0.56, 0.66)	151.925 4	47.278 8	2.273 9
	[0.66, 0.74)	239.892 3	29.977 2	0.913 1
	[0.74, 0.84]	313.504 2	16.940 7	0.394 8
	＞0.84	228.259 8	2.498 4	0.080 0

下载: 导出CSV

表 3 汶川地震滑坡评价因子间方差膨胀因子及容差

Table 3. Variance inflation factors and tolerances among evaluation factors of Wenchuan earthquake-induced landslide

评价因子	共线性统计量
评价因子	容差(TOL)	方差膨胀因子(VIF)
高程	0.351	2.845
坡度	0.966	1.035
坡向	0.988	1.012
曲率	0.991	1.009
地层岩性	0.902	1.109
距断层距离	0.348	2.872
距水系距离	0.477	2.095
距道路距离	0.301	3.320
PGV	0.202	4.945
PGA	0.160	6.239

下载: 导出CSV

表 4 评价模型参数

Table 4. Parameters setting of evaluation models

模型类别	参数设置
模型类别	决策树数目	分类最少样本数	分支最少样本数	最大深度	学习率	分类标准
FR-GBDT	131	96	31	7	0.2	/
FR-RF	248	2	1	18	/	Gini
GBDT	131	96	31	7	0.2	/
RF	248	2	1	18	/	Gini
注：Gini为模型分类的分裂标准

下载: 导出CSV

表 5 机器学习模型的易发性评价等级的统计结果

Table 5. Statistical results of susceptibility rating for machine learning models

模型	易发性等级	发生滑坡栅格数	分级栅格数	占总滑坡比例/%	占总栅格比例/%	滑坡比率
GBDT	极低	7 329	473 643	5.733 5	50.234 3	0.114 1
	低	14 842	120 674	11.611 0	12.798 6	0.907 2
	中	26 398	116 494	20.651 3	12.355 3	1.671 5
	高	37 528	118 579	29.358 4	12.576 4	2.334 4
	极高	41 730	113 477	32.645 7	12.035 3	2.712 5
RF	极低	2 701	392 808	2.113 0	41.661 0	0.050 7
	低	10 782	137 679	8.434 8	14.602 2	0.577 6
	中	25 973	137 478	20.318 9	14.580 8	1.393 5
	高	40 295	137 484	31.523 1	14.581 5	2.161 9
	极高	48 076	137 418	37.610 2	14.574 5	2.580 6
FR-GBDT	极低	6 419	507 743	5.021 6	53.851 0	0.093 3
	低	11 281	97 685	8.825 2	10.360 4	0.851 8
	中	15 522	77 212	12.143 0	8.189 1	1.482 8
	高	24 940	90 245	19.510 7	9.571 3	2.038 5
	极高	69 665	169 982	54.499 4	18.028 2	3.023 0
FR-RF	极低	134	433 468	0.104 8	45.973 4	0.002 3
	低	4 494	128 788	3.515 7	13.659 2	0.257 4
	中	19 605	125 002	15.337 1	13.257 6	1.156 9
	高	36 206	117 068	28.324 2	12.416 2	2.281 2
	极高	67 388	138 541	52.718 1	14.693 6	3.587 8

下载: 导出CSV

表 6 机器学习模型评价指标

Table 6. Evaluation indicators of machine learning models

模型类别	评价指标
模型类别	训练时间/min	预测准确度/%	受试者曲线下面积(AUC)
RF	81	82.43	0.901
GBDT	281	77.90	0.856
FR-RF	71	85.82	0.943
FR-GBDT	287	83.66	0.926

下载: 导出CSV

参考文献(49)

[1]	杨迁, 王雁林, 马园园. 2001-2019年中国地质灾害分布规律及引发因素分析[J]. 地质灾害与环境保护, 2020, 31(4): 43-48. doi: 10.3969/j.issn.1006-4362.2020.04.008 YANG Q, WANG Y L, MA Y Y. Distribution rule and influencing factors of geological disasters from 2001 to 2019 in China[J]. Journal of Geological Hazards and Environment Preservation, 2020, 31(4): 43-48. (in Chinese with English abstract) doi: 10.3969/j.issn.1006-4362.2020.04.008
[2]	李宏杰, 戴福初, 许冲. 地震滑坡研究现状综述[C]//佚名. 2011年AASRI智能信息技术应用学会论文集. [出版地不详]: [出版社不详], 2011: 172-179. LI H J, DAI F C, XU C. A review of the research on earthquake-induced landslides[C]//Anon. 2011 AASRI Conference on Applied Information Technology (AASRI-AIT 2011). [S. l.]: [s. n.], 2011: 172-179. (in Chinese with English abstract)
[3]	唐辉明. 重大滑坡预测预报研究进展与展望[J]. 地质科技通报, 2022, 41(6): 1-13. doi: 10.19509/j.cnki.dzkq.2022.0203 TANG H M. Advance and prospects of major landslides prediction and forecasting[J]. Bulletin of Geological Science and Technology, 2022, 41(6): 1-13. (in Chinese with English abstract) doi: 10.19509/j.cnki.dzkq.2022.0203
[4]	王兰民, 郭安宁, 王平, 等. 1920年海原大地震震害特征与启示[J]. 城市与减灾, 2020(6): 43-53. doi: 10.3969/j.issn.1671-0495.2020.06.007 WANG L M, GUO A N, WANG P, et al. Characteristics and revelation of great Haiyuan earthquake disaster in 1920[J]. City and Disaster Reduction, 2020(6): 43-53. (in Chinese with English abstract) doi: 10.3969/j.issn.1671-0495.2020.06.007
[5]	许冲, 徐锡伟, 吴熙彦, 等. 2008年汶川地震滑坡详细编目及其空间分布规律分析[J]. 工程地质学报, 2013, 21(1): 25-44. doi: 10.3969/j.issn.1004-9665.2013.01.004 XU C, XU X W, WU X Y, et al. Detailed catalog of landslides triggered by the 2008 Wenchuan earthquake and statistical analyses of their spatial distribution[J]. Journal of Engineering Geology, 2013, 21(1): 25-44. (in Chinese with English abstract) doi: 10.3969/j.issn.1004-9665.2013.01.004
[6]	王欣, 方成勇, 唐小川, 等. 泸定M_s 6.8地震诱发滑坡应急评价研究[J]. 武汉大学学报(信息科学版), 2023, 48(1): 25-35. https://www.cnki.com.cn/Article/CJFDTOTAL-WHCH202301003.htm WANG X, FANG C Y, TANG X C, et al. Research on emergency evaluation of landslides induced by the Luding M_s 6.8 earthquake[J]. Geomatics and Information Science of Wuhan University, 2023, 48(1): 25-35. (in Chinese with English abstract) https://www.cnki.com.cn/Article/CJFDTOTAL-WHCH202301003.htm
[7]	XIANG Z L, DOU J, YUNUS A P, et al. Vegetation-landslide nexus and topographic changes post the 2004 M_w 6.6 Chuetsu earthquake[J]. CATENA, 2023, 223: 106946. doi: 10.1016/j.catena.2023.106946
[8]	HUANG Y, ZHAO L. Review on landslide susceptibility mapping using support vector machines[J]. CATENA, 2018, 165: 520-529. doi: 10.1016/j.catena.2018.03.003
[9]	MERGHADI A, YUNUS A P, DOU J, et al. Machine learning methods for landslide susceptibility studies: A comparative overview of algorithm performance[J]. Earth-Science Reviews, 2020, 207: 103225. doi: 10.1016/j.earscirev.2020.103225
[10]	ADO M, AMITAB K, MAJI A K, et al. Landslide susceptibility mapping using machine learning: A literature survey[J]. Remote Sensing, 2022, 14(13): 3029. doi: 10.3390/rs14133029
[11]	冷伏海, 周秋菊, 杨帆, 等. 2020研究前沿[R]. 北京: 中国科学院科技战略咨询研究院, 2020. LENG F H, ZHOU Q J, YANG F, et al. 2020 research fronts[R]. Beijing: Institutes of Science and Development, Chinese Academy of Sciences, 2020.
[12]	BAI S B, WANG J, LÜ G N, et al. GIS-based logistic regression for landslide susceptibility mapping of the Zhongxian segment in the Three Gorges area, China[J]. Geomorphology, 2010, 115(1/2): 23-31.
[13]	CHEN W, XIE X S, WANG J L, et al. A comparative study of logistic model tree, random forest, and classification and regression tree models for spatial prediction of landslide susceptibility[J]. CATENA, 2017, 151: 147-160. doi: 10.1016/j.catena.2016.11.032
[14]	DOU J, YAMAGISHI H, POURGHASEMI H R, et al. An integrated artificial neural network model for the landslide susceptibility assessment of Osado Island, Japan[J]. Natural Hazards, 2015, 78(3): 1749-1776. doi: 10.1007/s11069-015-1799-2
[15]	MA Z J, MEI G. Deep learning for geological hazards analysis: Data, models, applications, and opportunities[J]. Earth-Science Reviews, 2021, 223: 103858. doi: 10.1016/j.earscirev.2021.103858
[16]	窦杰, 向子林, 许强, 等. 机器学习在滑坡智能防灾减灾中的应用与发展趋势[J]. 地球科学, 2023, 48(5): 1657-1674. https://www.cnki.com.cn/Article/CJFDTOTAL-DQKX202305001.htm DOU J, XIANG Z L, XU Q, et al. Application and development trend of machine learning in landslide intelligent disaster prevention and mitigation[J]. Earth Science, 2023, 48(5): 1657-1674. (in Chinese with English abstract) https://www.cnki.com.cn/Article/CJFDTOTAL-DQKX202305001.htm
[17]	黄发明, 胡松雁, 闫学涯, 等. 基于机器学习的滑坡易发性预测建模及其主控因子识别[J]. 地质科技通报, 2022, 41(2): 79-90. doi: 10.19509/j.cnki.dzkq.2021.0087 HUANG F M, HU S Y, YAN X Y, et al. Landslide susceptibility prediction and identification of its main environmental factors based on machine learning models[J]. Bulletin of Geological Science and Technology, 2022, 41(2): 79-90. (in Chinese with English abstract) doi: 10.19509/j.cnki.dzkq.2021.0087
[18]	杨灿, 刘磊磊, 张遗立, 等. 基于贝叶斯优化机器学习超参数的滑坡易发性评价[J]. 地质科技通报, 2022, 41(2): 228-238. doi: 10.19509/j.cnki.dzkq.2022.0059 YANG C, LIU L L, ZHANG Y L, et al. Machine learning based on landslide susceptibility assessment with Bayesian optimized the hyperparameters[J]. Bulletin of Geological Science and Technology, 2022, 41(2): 228-238. (in Chinese with English abstract) doi: 10.19509/j.cnki.dzkq.2022.0059
[19]	LUO W Q, DOU J, FU Y H, et al. A novel hybrid LMD-ETS-TCN approach for predicting landslide displacement based on GPS time series analysis[J]. Remote Sensing, 2022, 15(1): 229. doi: 10.3390/rs15010229
[20]	NI W D, ZHAO L Y, ZHANG L L, et al. Coupling progressive deep learning with the AdaBoost framework for landslide displacement rate prediction in the Baihetan Dam Reservoir, China[J]. Remote Sensing, 2023, 15(9): 2296. doi: 10.3390/rs15092296
[21]	DOU J, YUNUS A P, BUI D T, et al. Improved landslide assessment using support vector machine with bagging, boosting, and stacking ensemble machine learning framework in a mountainous watershed, Japan[J]. Landslides, 2020, 17(3): 641-658. doi: 10.1007/s10346-019-01286-5
[22]	DOU J, YUNUS A P, TIEN BUI D, et al. Assessment of advanced random forest and decision tree algorithms for modeling rainfall-induced landslide susceptibility in the Izu-Oshima Volcanic Island, Japan[J]. Science of the Total Environment, 2019, 662: 332-346. doi: 10.1016/j.scitotenv.2019.01.221
[23]	UMAR Z, PRADHAN B, AHMAD A, et al. Earthquake induced landslide susceptibility mapping using an integrated ensemble frequency ratio and logistic regression models in West Sumatera Province, Indonesia[J]. CATENA, 2014, 118: 124-135. doi: 10.1016/j.catena.2014.02.005
[24]	ZHU A X, MIAO Y M, YANG L, et al. Comparison of the presence-only method and presence-absence method in landslide susceptibility mapping[J]. CATENA, 2018, 171: 222-233. doi: 10.1016/j.catena.2018.07.012
[25]	CHOI J, OH H J, LEE H J, et al. Combining landslide susceptibility maps obtained from frequency ratio, logistic regression, and artificial neural network models using ASTER images and GIS[J]. Engineering Geology, 2012, 124: 12-23. doi: 10.1016/j.enggeo.2011.09.011
[26]	DOU J, TIEN BUI D, YUNUS A P, et al. Optimization of causative factors for landslide susceptibility evaluation using remote sensing and GIS data in parts of Niigata, Japan[J]. PLoS One, 2015, 10(7): e0133262. doi: 10.1371/journal.pone.0133262
[27]	周晓亭, 黄发明, 吴伟成, 等. 基于耦合信息量法选择负样本的区域滑坡易发性预测[J]. 工程科学与技术, 2022, 54(3): 25-35. https://www.cnki.com.cn/Article/CJFDTOTAL-SCLH202203003.htm ZHOU X T, HUANG F M, WU W C, et al. Regional landslide susceptibility prediction based on negative sample selected by coupling information value method[J]. Advanced Engineering Sciences, 2022, 54(3): 25-35. (in Chinese with English abstract) https://www.cnki.com.cn/Article/CJFDTOTAL-SCLH202203003.htm
[28]	LIU L L, ZHANG Y L, XIAO T, et al. A frequency ratio-based sampling strategy for landslide susceptibility assessment[J]. Bulletin of Engineering Geology and the Environment, 2022, 81(9): 360. doi: 10.1007/s10064-022-02836-3
[29]	殷跃平. 汶川八级地震地质灾害研究[J]. 工程地质学报, 2008, 16(4): 433-444. doi: 10.3969/j.issn.1004-9665.2008.04.001 YIN Y P. Researches on the geo-hazards triggered by Wenchuan earthquake, Sichuan[J]. Journal of Engineering Geology, 2008, 16(4): 433-444. (in Chinese with English abstract) doi: 10.3969/j.issn.1004-9665.2008.04.001
[30]	FAN X M, SCARINGI G, DOMÈNECH G, et al. Two multi-temporal datasets that track the enhanced landsliding after the 2008 Wenchuan earthquake[J]. Earth System Science Data, 2019, 11(1): 35-55. doi: 10.5194/essd-11-35-2019
[31]	OZDEMIR A, ALTURAL T. A comparative study of frequency ratio, weights of evidence and logistic regression methods for landslide susceptibility mapping: Sultan Mountains, SW Turkey[J]. Journal of Asian Earth Sciences, 2013, 64: 180-197. doi: 10.1016/j.jseaes.2012.12.014
[32]	LIU R, LI L Y, PIRASTEH S, et al. The performance quality of LR, SVM, and RF for earthquake-induced landslides susceptibility mapping incorporating remote sensing imagery[J]. Arabian Journal of Geosciences, 2021, 14(4): 259. doi: 10.1007/s12517-021-06573-x
[33]	MEZAAL M R, PRADHAN B. An improved algorithm for identifying shallow and deep-seated landslides in dense tropical forest from airborne laser scanning data[J]. CATENA, 2018, 167: 147-159. doi: 10.1016/j.catena.2018.04.038
[34]	SCHONLAU M, ZOU R Y. The random forest algorithm for statistical learning[J]. The Stata Journal: Promoting Communications on Statistics and Stata, 2020, 20(1): 3-29. doi: 10.1177/1536867X20909688
[35]	FRIEDMAN J H. Stochastic gradient boosting[J]. Computational Statistics & Data Analysis, 2002, 38(4): 367-378.
[36]	王恒, 姜亚楠, 张欣, 等. 基于梯度提升算法的岩性识别方法[J]. 吉林大学学报(地球科学版), 2021, 51(3): 940-950. https://www.cnki.com.cn/Article/CJFDTOTAL-CCDZ202103026.htm WANG H, JIANG Y N, ZHANG X, et al. Lithology identification method based on gradient boosting algorithm[J]. Journal of Jilin University (Earth Science Edition), 2021, 51(3): 940-950. (in Chinese with English abstract) https://www.cnki.com.cn/Article/CJFDTOTAL-CCDZ202103026.htm
[37]	李永威, 徐林荣, 张亮亮, 等. 强震山区地震诱发滑坡发育规律与易发性评估[J]. 地球科学, 2023, 48(5): 1960-1976. https://www.cnki.com.cn/Article/CJFDTOTAL-DQKX202305020.htm LI Y W, XU L R, ZHANG L L, et al. Study on development patterns and susceptibility evaluation of coseismic landslides within mountainous regions influenced by strong earthquakes[J]. Earth Science, 2023, 48(5): 1960-1976. (in Chinese with English abstract) https://www.cnki.com.cn/Article/CJFDTOTAL-DQKX202305020.htm
[38]	蔡生来. 安徽怀宁县崩塌、滑坡和泥石流孕灾地质条件特征研究[J]. 安徽地质, 2021, 31(3): 265-269. doi: 10.3969/j.issn.1005-6157.2021.03.017 CAI S L. A study on the features of the geological conditions inducing hazards of collapse, landslide and debris flow in Huaining County, Anhui Province[J]. Geology of Anhui, 2021, 31(3): 265-269. (in Chinese with English abstract) doi: 10.3969/j.issn.1005-6157.2021.03.017
[39]	周毅, 丁明涛, 黄涛, 等. 芦山县滑坡灾害影响因素的空间分异性[J]. 中国地质调查, 2022, 9(4): 45-55. https://www.cnki.com.cn/Article/CJFDTOTAL-DZDC202204006.htm ZHOU Y, DING M T, HUANG T, et al. Spatial heterogeneity of influencing factors of landslide disasters in Lushan County[J]. Geological Survey of China, 2022, 9(4): 45-55. (in Chinese with English abstract) https://www.cnki.com.cn/Article/CJFDTOTAL-DZDC202204006.htm
[40]	柳万里. 含泥岩类夹层巴东组斜坡工程地质特性及其孕滑机理研究[D]. 武汉: 中国地质大学(武汉), 2022. LIU W L. Study on engineering geological characteristics and pregnant sliding mechanism of Badong Formation slope with mudstone interlayer[D]. Wuhan: China University of Geosciences (Wuhan), 2022. (in Chinese with English abstract)
[41]	FAN X M, SCARINGI G, XU Q, et al. Coseismic landslides triggered by the 8th August 2017 M_s 7.0 Jiuzhaigou earthquake (Sichuan, China): Factors controlling their spatial distribution and implications for the seismogenic blind fault identification[J]. Landslides, 2018, 15(5): 967-983. doi: 10.1007/s10346-018-0960-x
[42]	XU C, XU X W, YAO X, et al. Three (nearly) complete inventories of landslides triggered by the May 12, 2008 Wenchuan M_w 7.9 earthquake of China and their spatial distribution statistical analysis[J]. Landslides, 2014, 11(3): 441-461. doi: 10.1007/s10346-013-0404-6
[43]	PHAM B T, PRAKASH I, SINGH S K, et al. Landslide susceptibility modeling using Reduced Error Pruning Trees and different ensemble techniques: Hybrid machine learning approaches[J]. CATENA, 2019, 175: 203-218. doi: 10.1016/j.catena.2018.12.018
[44]	陈飞, 蔡超, 李小双, 等. 基于信息量与神经网络模型的滑坡易发性评价[J]. 岩石力学与工程学报, 2020, 39(增刊1): 2859-2870. https://www.cnki.com.cn/Article/CJFDTOTAL-YSLX2020S1027.htm CHEN F, CAI C, LI X S, et al. Evaluation of landslide susceptibility based on information valueme and neural network model[J]. Chinese Journal of Rock Mechanics and Engineering, 2020, 39(S1): 2859-2870. (in Chinese with English abstract) https://www.cnki.com.cn/Article/CJFDTOTAL-YSLX2020S1027.htm
[45]	王世宝, 庄建琦, 郑佳, 等. 基于深度学习的CZ铁路康定-理塘段滑坡易发性评价[J]. 工程地质学报, 2022, 30(3): 908-919. https://www.cnki.com.cn/Article/CJFDTOTAL-GCDZ202203028.htm WANG S B, ZHUANG J Q, ZHENG J, et al. Landslide susceptibility evaluation based on deep learning along Kangding-Litang section of CZ railway[J]. Journal of Engineering Geology, 2022, 30(3): 908-919. (in Chinese with English abstract) https://www.cnki.com.cn/Article/CJFDTOTAL-GCDZ202203028.htm
[46]	CHANG K T, MERGHADI A, YUNUS A P, et al. Evaluating scale effects of topographic variables in landslide susceptibility models using GIS-based machine learning techniques[J]. Scientific Reports, 2019, 9(1): 12296. doi: 10.1038/s41598-019-48773-2
[47]	郭子正, 殷坤龙, 黄发明, 等. 基于滑坡分类和加权频率比模型的滑坡易发性评价[J]. 岩石力学与工程学报, 2019, 38(2): 287-300. https://www.cnki.com.cn/Article/CJFDTOTAL-YSLX201902007.htm GUO Z Z, YIN K L, HUANG F M, et al. Evaluation of landslide susceptibility based on landslide classification and weighted frequency ratio model[J]. Chinese Journal of Rock Mechanics and Engineering, 2019, 38(2): 287-300. (in Chinese with English abstract) https://www.cnki.com.cn/Article/CJFDTOTAL-YSLX201902007.htm
[48]	FRATTINI P, CROSTA G, CARRARA A. Techniques for evaluating the performance of landslide susceptibility models[J]. Engineering Geology, 2010, 111(1/4): 62-72.
[49]	崔玉龙, 刘爱娟. 区域边坡地震危险性评价理论研究进展[J]. 地震工程学报, 2022, 44(3): 518-526. https://www.cnki.com.cn/Article/CJFDTOTAL-ZBDZ202203003.htm CUI Y L, LIU A J. Advances in the theory of seismic hazard assessment of regional slopes[J]. China Earthquake Engineering Journal, 2022, 44(3): 518-526. (in Chinese with English abstract) https://www.cnki.com.cn/Article/CJFDTOTAL-ZBDZ202203003.htm