Application of different machine learning models in landslide susceptibility assessment in Badong County, Hubei province
-
摘要:
中国是世界上发生滑坡灾害最频繁的国家之一,滑坡易发性评价有助于防灾减灾工作。由于不同机器学习模型在不同区域的适配程度不同,为更好开展湖北省巴东县的滑坡灾害防治工作,选取坡度、坡向、曲率、起伏度、地层、覆盖层、NDVI、道路密度、水系密度、斜坡结构10个影响因子,采用逻辑回归(LR)、支持向量机(SVM)、多层感知机(MLP)和随机森林(RF)四种模型进行滑坡易发性评价。并通过受试者工作特征曲线(ROC)、均方误差与决定系数等指标、滑坡-研究区占比三种评价方式用于评价模型精度。实验结果表明,不同模型在不同评价方式中存在差异,但总体而言,RF模型精度最高且绘制出的易发性分区图更合理。四个模型绘制的易发性区域分布图相似,极高易发区和高易发区主要分布于南边沿江地区,西南沿岸的官渡口镇、焦家湾村等附近地区表现出较高易发性,该评价结果可以为巴东县的滑坡治理提供参考。
Abstract:China is one of the countries most frequently affected by landslide disasters in the world, making landslide susceptibility assessment crucial for effective disaster prevention and mitigation. Due to variations in the adaptability of different machine learning models in different regions, in order to better carry out landslide disaster prevention and control work in Badong County, Hubei Province, ten influencing factors including slope gradient, slope direction, curvature, degree of undulation, stratigraphy, overburden, NDVI, road density, water system density, and slope structure were selected. Four different models, including Logistic Regression (LR), Support Vector Machine (SVM), Multilayer Perceptron (MLP), and Random Forest (RF), were used for landslide susceptibility evaluation. Three evaluation methods were used to assess the accuracy of the model: Receiver Operating Characteristic (ROC) curves, mean square error, determination coefficient, and the ratio of landslide to study area. The experimental results show that there are differences among the models in different evaluation methods. Overall, the RF model exhibits the highest accuracy and generates more reasonable susceptibility zoning maps. The susceptibility distribution maps generated by the four models are similar, with high and very high susceptibility areas predominantly located in the southern riverside area. Areas near Guandukou Town and Jiaojiawan Village along the southwest coast exhibit relatively high susceptibility. The assessment results can provide reference for landslide control in Badong County.
-
0. 引言
湖北省三峡库区是全国滑坡灾害发生最频繁的地区之一,库内现共有滑坡4664个,其中674个有明显变形特征[1]。巴东县是三峡库区地质灾害最严重的地区之一,也是全国地质灾害研究的重点地区[2],开展滑坡易发性评价可为滑坡灾害的预测预警防治提供技术支持,是现阶段滑坡治理的迫切任务。滑坡易发性评价中斜坡单元的划分是重要环节之一,目前划分的方法主要有:水文分析法、地表曲率分水岭法、r.slopeunits方法等[3 − 5]。地质灾害治理得益于过去十多年内人工智能及大数据技术的发展[6],现机器学习在灾害防治方面得到广泛应用并做出很大的贡献,如王俊得、刘宝生、牟家琦等[7 − 11]采用各种模型进行灾害易发性评价,常见的评价模型有:证据权法(Weight of Evidence,WOE)、随机森林(Random Forest,RF)、支持向量机(Support Vector Machine,SVM)、逻辑回归(Logical Regression,LR)、神经网络(Neural Networks,NN)等[12 − 15]。各种模型都有缺陷,如机器学习没有解释个别特征对结果的影响,也没考虑它们的相互依赖性,而深度神经网络缺乏可解释性[16]。所有的模型都有其特定的性质和缺点,每个模型的性能根据输入数据、模型的结构和准确性而变化[17]。因此,单一的模型不适用于所有区域,不同地区的同种预测模型产生的结果存在差异,在进行灾害易发性研究时,需根据研究地区的实际条件进行模型的适用性选择。为更好的开展巴东县滑坡防治工作,选取多种机器学习模型进行滑坡易发性评价并对比各个模型精度,找出更适应巴东县地质情况的模型,从而提高滑坡易发性评价的针对性,为巴东县滑坡防治工作提供参考。
针对上述问题,本文通过支持向量机(SVM)、逻辑回归(LR)、多层感知机(MLP)、随机森林(RF)四种机器学习模型对巴东县进行滑坡易发性评价,获取易发值后将研究区域划分为极高易发区、高易发区、中易发区、低易发区四个区域。并通过受试者工作特征曲线(Receiver Operating Characteristic,ROC)、数理统计等方法对模型精度进行评估。
1. 研究区概况
本文研究区位于湖北省巴东县,东至东瀼口镇,西至神农小区,北至谭家屋场,南邻长江,东经110°19′15″~110°22′02″,北纬31°03′00″~31°05′45″,总面积约10122333.4 m2。
巴东城区四水环绕、山城镶嵌、地表崎岖、山峦起伏、峡谷幽深、沟壑纵横,是典型的喀斯特地貌[18]。三峡库区四季分明,气候湿润,年均降雨量近1500 mm,水系交错纵横,植被覆盖率高达55%以上[19]。复杂的环境造成巴东县有许多制约发展和危害安全的潜在地质问题,如滑坡、泥石流等[20 − 23]。
研究区地势西南高东北低,滑坡密集发育,现共有25个滑坡,主要分布于沿江地区。研究区滑坡的高程分布为142.05 m~657.53 m,坡度分布为0°~63.27°。滑床地层岩性主要为泥质粉砂岩,滑体主要为碎石土。
2. 研究方法
2.1 指标及评价因子
网格单元难以反应出地貌和滑坡灾害的关系特征,而斜坡单元与塑造地貌的条件和过程相关,是滑坡发育的基本单元[24]。本文采用斜坡单元作为评价单元,采用水文分析法进行斜坡单元的划分。
由于滑坡区域常横跨多个斜坡单元,为精确描述滑坡面积,采用10 m×10 m的栅格单元,采用栅格数占比代替面积占比。
影响某地区滑坡发生的环境因素十分复杂。因此,很难确定哪些环境因素是最重要和必要的[25]。刘丽娜[26]等指出坡度、坡向、曲率是滑坡形成主控因子中的地形因子。李松林[27]等指出斜坡结构能表现出地形因子的关系,是滑坡发育的关键影响因子之一。石菊松[28]等指出地表水的冲蚀作用对滑坡有突出影响,水系的密集程度是滑坡发生的影响因子之一。Wang等[29]指出断层边的滑坡运动会相互造成影响,断层距离是常见的滑坡影响因子。但在本研究区内不存在活动断层,断层距离对滑坡的影响几乎可以忽略,因此不作为模型训练的参数。郭慧娟等[30]指出人类工程活动中修建道路会对两侧的岩土造成影响,诱导滑坡的发生。章昱等[31]指出植被发育程度直接影响斜坡稳定性。高程体现斜坡的势能和内部应力[32],而起伏度为高程变化程度,体现了一定范围内势能差,因而陈刚[33]等指出起伏度是滑坡发生的指标因子。张玺国[34]等研究发现地层岩性在影响地质灾害中重要性达1%~10%。皱浩[35]等指出覆盖层孕育着斜坡浅表层破坏,是使斜坡从稳定态向不稳定态变化的主要原因,且覆盖层的空间变化圈定滑坡发育的边界。基于前人的研究,结合研究区的地质状况,本文选取坡度、坡向、曲率、起伏度、地层岩性、覆盖层、NDVI、道路密度、水系密度、斜坡结构10个可量化提取的影响因子(图1)。
2.2 机器学习模型
机器学习是滑坡易发性评价常用的模型,本文采用了SVM、LR、MLP和RF四种模型对滑坡易发性进行评价。
(1)支持向量机
SVM在解决小样本、非线性和高维模式识别问题中有特有的优势[36]。对于样本(xi , yi),i=1, 2, 3···n,寻找一个超平面:
$$ \omega x + b = 0 $$ (1) 使得支持向量到超平面(1)的距离最大化,公式为:
$$ \max d = \frac{{|\omega x + b|}}{{\left\| \omega \right\|}} $$ (2) 引入拉格朗日函数,将问题向高纬度投影,并使
$ \omega $ 和$ b $ 偏导数为0,可将问题转化为:$$ \mathop {\min }\limits_{\omega ,b} L(\omega ,b,\alpha ) = \sum\limits_{j = 1}^n {{\alpha _j}} - \frac{1}{2}\sum\limits_{i = 1}^n {\sum\limits_{j = 1}^n {{\alpha _i}{\alpha _j}{y_i}{y_j}({x_i} * {x_j})} } $$ (3) 式中,
$ \alpha = {({\alpha _1},{\alpha _2}\cdots{\alpha _n})^T} $ 可通过SMO等算法来获取。(2)逻辑回归
LR是一种多变量分析模型,可以根据一组预测变量的值来预测特征或结果的存在与否[37]。传统的多元线性回归方程:
$$ y = {\beta _0} + \sum\limits_{i = 1}^n {{x_i}{\beta _i}} $$ (4) LR使用的联系函数为对数几率函数:
$$ \textit{z} = {\mathrm{In}}\frac{y}{{1 - y}} $$ (5) 将(4)中的y换成z带入式(5)中,可得LR的数学原型:
$$ y = \frac{1}{{1 + {e^{ - ({\beta _0} + {\beta _1}{x_1} + \cdots + {\beta _n}{x_n})}}}} $$ (6) 对式(6)进行变形后,可得LR常用数学公式:
$$ y = \frac{{{e^{({\beta _0} + {\beta _1}{x_1} + \cdots + {\beta _n}{x_n})}}}}{{1 + {e^{({\beta _0} + {\beta _1}{x_1} + \cdots + {\beta _n}{x_n})}}}} $$ (7) 式中,
$ \beta =({\beta }_{0},{\beta }_{1},\mathrm{\cdots},{\beta }_{n}) $ 可通过极大似然估计法求解。(3)多层感知机
MLP是最常见的全连接人工神经网络(ANN)[38],MLP由输入层、隐藏层、输出层三层构成[39]。输入层节点分配权重构建加权线性方程作为下一层的输入,隐藏层节点值公式为:
$$ {y_{}} = f(\omega x + b) $$ (8) 式中:
$ y $ 、$ \omega $ 、$ x $ ——向量,隐藏层的数量便是其嵌套数量。 经过向前阶段后,本文通过Sigmoid激 活函数:$$ Sigmoid(x) = \frac{1}{{1 + \exp ( - x)}} $$ (9) 及其偏导数进行反向传播,不断更新网络的权值
$ \omega $ ,使网络缓慢收敛。(4)随机森林
RF是一种集成学习模型(Ensemble Learning),其算法是是一种bagging算法,它由多颗决策树组成,并且由大多数决策树分类结果来决定样本类型[40]。在回归分析中,则是取各个决策树的回归结果的平均值作为结果。
2.3 研究思路
本文研究思路(图2)分为数据获取、模型训练、易发值处理三部分。
数据处理部分,本文所用到的数据均来自资助项目,从收集到的数据中,选取2.1节中的10个指标用于训练。
模型训练部分,通过选取的特征参数,生成样本集,通过测试取得较优的模型参数。其中,SVM通过AUC面积确定参数C的较优值;RF通过均方误差确定较优的决策树数量;MLP通过多次测试选取较优的层数和节点数。选定较优参数进行模型训练,以斜坡单元为评价单元,进行研究区的滑坡易发性评价,并通过ROC曲线和数理统计指标用于评价模型精度。
易发值处理部分,获取模型预测的易发值后,剔除异常的易发值,将易发值归一化后得到处理后的易发值。将处理后的易发值通过自然间断点重分类为4类并生成易发分区图,根据易发性分区图和原有滑坡数据用于数据分析评价模型精度。最后,将ROC曲线、数理统计、数据分析三个部分得出的评价结果用于进行综合评价,得出四个模型最终精度评价结果。
2.4 数据处理
数据处理包含剔除异常值和数据归一化两个部分。本文采用的归一化方法为最大-最小标准化,其公式为:
$$ {x_{i,new}} = \frac{{{x_i} - \max (x)}}{{\max (x) - \min (x)}} $$ (10) 式中:xi,new——第i个数据归一化后的值;
max(x)——x数据集中的最大值;
min(x)——x数据集中的最小值。
在实际应用在中,常出现极少数预测值偏离数据中心,呈现极大或极小的现象,造成式(10)中的分母过大,归一化后大部分预测值极小的情况。本文通过异常斜坡附近的斜坡单元易发性值对其预测,假设该斜坡单元的序号为i,其预测值为:
$$ \left|Valu{e}_{i}\right|=\left\{\begin{array}{l}\frac{{\displaystyle \sum _{n=i+1}^{i+k}\left|Valu{e}_{n}\right|}}{k}\text{,}i < k\\ \frac{{\displaystyle \sum _{n=i-k,n\ne i}^{i+k}\left|Valu{e}_{n}\right|}}{2k}\text{,}k < =i < =\mathrm{max}si\textit{z}e-k\\ \frac{{\displaystyle \sum _{n=i-k}^{i-1}\left|Valu{e}_{n}\right|}}{k}\text{,}i > \mathrm{max}si\textit{z}e-k\end{array}\right. $$ (11) 式中:Value——斜坡单元的易发性值;
maxsize——斜坡单元的最大序号;
i,n——斜坡单元序号;
k——选取邻近斜坡单元的个数。
式(11)表示,当斜坡周围的斜坡单元数量少于2k时,选取前或后k个斜坡单元的绝对值均值作为预测值绝对值;若多于2k个,则取前k个和后k个斜坡单元绝对值均值作为预测值绝对值。获得绝对值后,原易发性值的正负作为更新易发值的正负,公式为:
$$ Valu{e}_{i}=\left\{\begin{aligned} &\left|Valu{e}_{i}\right|,\;\;if\;(Valu{e}_{old} > =0)\\ &-\left|Valu{e}_{i}\right|,\;\;if\;(Valu{e}_{old} < 0)\end{aligned} \right.$$ (12) 式中:Valuei——异常斜坡单元更新后的易发值;
Valueold——异常斜坡单元的原易发值;
if——条件语句,括号内为真则选择该项。
3. 研究结果与讨论
本文于灾害区和非灾害区内创建随机点,选取相同数量的灾害点和非灾害点作为样本,样本数量共计 800个。将样本数量划分为70%训练集,30%测试集两部分。预测区域划分为245个斜坡单元,通过ArcGIS将斜坡单元与评价因子相结合并转化为预测点集。
本文通过ROC曲线、数理统计指标、滑坡-研究区占比用于模型评估。其中,滑坡-研究区占比的评价依据是:通过已知的滑坡区域(灾害区)获取灾害点数据,由于所选区域本身就是滑坡区域,即高危险区域,因此训练出来的模型如果表现为高易发值,就与实际情况比较符合,对应模型的模拟效果也就比较好。采用滑坡内部分区占比和滑坡占据研究区不同分区的占比两种占比用于评价。其中,滑坡内部分区应大部分比例为极高易发区或高易发区,低易发区比例越少越符合现实情况;滑坡占据研究区各分区占比中,滑坡区域在极高易发区或高易发区的占比越高,低易发区的占比越少视作更加符合现实情况。
3.1 基于SVM模型的易发性评价结果
SVM对参数调节和核函数的选择敏感[41],为了使模型更加精确,在参数C∈[1,500]下绘制其ROC曲线,计算AUC面积,AUC面积代表模型精确度,面积越接近1,模型精度越高[42]。绘制AUC面积与C值关系曲线(图3)。
由图3可见,随着C值的增大,模型在训练集和测试集的精确度越来越高,但测试集精确度在C=200后增大不再显著,为防止模型过拟合与减少复杂度,选择参数C的值为200进行训练。将训练后的模型用于对斜坡单元进行预测,预测的易发性值进行剔除异常值、归一化后通过ArcGIS进行重分类得到易发性评价分区图(图4)。
计算滑坡内部分区的栅格数占比(图5)用于模型精确度分析。滑坡区域中极高易发区和高易发区的占比为59.36%,超过一半的滑坡区域为高易发值区域,但在中低易发区的比例达到40.64%,可见预测结果与真实值稍有偏差。
3.2 基于LR模型的易发性评价结果
LR实际上是检查独立因素对二元结果的贡献的系统且强大的模型[43]。通过训练模型获得独立因素对分类结果的贡献值,即式(7)中的(β1, β2, ···, βn),式中的自变量(x1, x2, ···, xn)为斜坡易发性评级因子(图1)。根据式(7)通过ArcGIS的“栅格计算器”功能,可计算得到斜坡易发性分区图(图6)。
计算滑坡内部分区的栅格数占比(图7)用于模型精确度分析。滑坡区域中极高易发区占比46.13%,高易发区占比32.60%,两区占比共计78.73%,中低易发区占比共计21. 27%。滑坡区域中大部分为高易发值区域,小部分为中低易发区,可见模型精度较高。
3.3 基于MLP模型易发性评价结果
MLP隐藏层的层数与节点数难以找到最佳的结果。在通过多次测试后,选择层数为5,节点数分别为160,125,90,75,40的结果较稳定良好。通过模型训练后进行易发值预测,最后对易发性图进行重分类可得易发性评价分区图(图8)。
计算滑坡内部分区的栅格数占比(图9)用于模型精确度分析。滑坡区域中极高易发区比例为
26.58%,高易发区占52.09%,两区共计78.67%,低易发区仅占3.31%,中低易发两区共占21.33%。滑坡区域中的中高易发区较高,低易发值只占小部分,由此可见模型精度较高。
3.4 基于RF模型的易发性评价结果
RF的调参是一个关键步骤,既要建立优良性能的模型,又要防止模型过拟合[44]。其中,最为重要的参数是森林中树木的数量,树木的数量越多模型的效果往往越好,但决策树数量过多会出现过拟合问题,因此需先寻找合适的决策树数量。模型精确度可通过训练集和测试集的均方误差体现。现绘制训练集和测试集的均方误差和树木数量关系曲线图(图10)。
由图10可见,在1至100的范围内,随着树木数量增加,模型精度不断提高,且测试集的均方误差趋近平稳,证明模型没有过拟合。在树木数量为40之后,精度变化不显著,继续增加树木数量会增大算法复杂度,因此可取树木数量为40的模型用于预测。将预测结果通过图1中的易发值处理流程后得到易发性分区图(图11)。
计算滑坡内部分区的栅格数占比(图12)用于模型精确度分析。滑坡区域中极高易发区占52.77%,高易发区占34.34%。滑坡区域中有87.11%的区域为高易发值区域,且低易发区占比低至3.19%,滑坡区域中高易发值区域占比接近90%,低易发值区域占比仅为12.89%,可见模型精度非常高。
3.5 结果讨论
本文通过四种模型对巴东县滑坡易发性进行了评价,为分析易发性分区评价结果,统计四种模型分区占比(表1),MLP的评价值结果在高易发区和中易发区的覆盖率较高,均在30%以上,而极高易发区显著低于其他模型;SVM的评价值较为保守,中低易发区比例相对较高,覆盖率共达64.08%;LR的各分区分布比例较为均匀,均在20%~30%之间;RF的低易发区比例较高,其他三个分区的比例相近。
表 1 各模型评价结果分区占比表Table 1. Proportion of Landslide Zone Assessment Results for Each Model极高/% 高/% 中/% 低/% LR 26.36 23.98 24.29 25.37 SVM 16.53 19.39 29.34 34.74 MLP 12.24 36.61 33.04 18.11 RF 20.16 25.58 20.75 33.51 观察各模型的易发性评价结果图(图4、6、8、11)的各分区分布情况,可见四种模型的评价结果相近,极高易发区和高易发区都大部分分布于西南的长江沿岸地区,部分区域沿着南岸分布,低易发区主要分布于西北区域。极高易发区和高易发区都主要分布于官渡口镇、焦家湾村等沿岸地区附近。相似的地区分布结果表明各模型对滑坡易发性评价的一致性和较高可信度。
通过各模型的ROC曲线(图13)对四种模型进行对比,ROC下的AUC可表示模型精确度。LR模型的AUC=0.74,SVM模型的AUC=0.76,MLP模型的AUC=0.73,RF模型的AUC=0.91。在AUC值方面,所有模型都取得良好的模型性能(AUC>0.70),可见四种模型都取得较好的效果。其中,RF对滑坡易发性的评价效果最佳,达到了0.90 以上,其他三种模型的评价效果相似,均分布在0.70~0.80之间。
考虑到不同模型的不确定性影响,进一步通过数理统计方法来判断四种回归模型中预测值和真实值的差异,常用指标为均方误差(Mean Square Error, MSE)、决定系数(Coefficient of Determination,CD)、平均绝对误差(Mean Absolute Error,MAE)和均方根误差(Root Mean Squared Error,RMSE),本文将通过上述四项指标评价模型在数据上的拟合程度。计算各模型的四项指标(图14),根据上述指标介绍的顺序排序,LR的各项指标为1.06、−0.06、0.53、1.03,SVM的各项指标为0.86、0.14、0.73、0.93,MLP的各项指标为2.60、−1.60、0.98、1.61,RF的各项指标为0.51、0.49、0.59、0.72。其中,CD值越接近1表明预测值和真实值的差异越小,MSE、MAE、RMSE值越
接近0表明预测值和真实值差异越小。由此可见,MLP的CD值和MSE值与其他模型的值偏差较多,预测结果和真实值差异性较大,而RF的预测结果优于其他三个模型。
在特殊情况下,模型预测的易发值过度偏大,造成极高易发区和高易发区范围比例过度偏高,滑坡内部分区占比的评价依据发生误判。为使滑坡-研究区占比的评价方式更加精确,统计滑坡区域占据研究区不同分区的占比(图15),即滑坡占据各易发区的比例。从滑坡占据研究区各分区的比例观察,训练的模型在灾害区应表现出高易发值,理想情况下灾害区域应分布在极高易发区,因此滑坡地区占据极高易发区的比例越高越符合现实情况,高易发区次之;滑坡区域若是出现在低易发值区域则不合理,因此占据低易发区的比例越高越不符合现实情况。总体而言,滑坡区域极大的占据RF预测结果中的极高易发区,占据比例接近70%,而LR的滑坡区域在高易发值区占比稍低,但SVM、LR和MLP的结果精度总体相近。由此可见,RF的预测结果最佳,SVM、LR和MLP的结果相近,在三种评价方法上都略低于RF模型。
4. 结论
(1) 四种模型的评价结果显示高易发值区主要分布在南边沿江地区,低易发值区主要分布于西北区域,西南沿岸的官渡口镇、焦家湾村等附近地区展现出较高易发性,相似的评价结果表明评价的一致性和较高的可信度。
(2) 实验表明,在AUC方面,RF有显著优势(AUC>0.9),其余三个模型相近(0.8>AUC>0.7),可见RF有着较优的模型精度;在预测值和实际值的近似度方面,MLP表现出较大偏差;在高易发值区域和实际滑坡区域的相似度方面,SVM的相似度最低,滑坡区域的中低易发区比例达到40.64%;LR的极高易发区中滑坡的占比为44.65%,略低于其他模型。可见不同模型在不同评价方式中存在差异,但总体可见,RF模型精度最高且绘制出的易发性分区图更合理。
(3) 通过不同机器学习模型在巴东城区的滑坡易发性评价应用研究,可以为三峡库区类似的重点城镇滑坡易发性评价工作提供参考范例,为三峡库区重点城镇的规划及可持续发展提供技术支持。
-
表 1 各模型评价结果分区占比表
Table 1 Proportion of Landslide Zone Assessment Results for Each Model
极高/% 高/% 中/% 低/% LR 26.36 23.98 24.29 25.37 SVM 16.53 19.39 29.34 34.74 MLP 12.24 36.61 33.04 18.11 RF 20.16 25.58 20.75 33.51 -
[1] 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(5):141 − 152. [WU Hongyang,ZHOU Chao,LIANG Xin,et al. Assessment of landslide susceptibility mapping based on XGBoost model:a case study of Yanshan Township[J]. The Chinese Journal of Geological Hazard and Control,2023,34(5):141 − 152. (in Chinese with English abstract)] WU Hongyang, ZHOU Chao, LIANG Xin, et al. Assessment of landslide susceptibility mapping based on XGBoost model: a case study of Yanshan Township[J]. The Chinese Journal of Geological Hazard and Control, 2023, 34(5): 141 − 152. (in Chinese with English abstract)
[2] 陈国金. 三峡库区巴东城区岸坡稳定性问题调查研究与风险控制[J]. 资源环境与工程,2023,37(3):292 − 300. [CHEN Guojin. Stability investigation and risk control of bank slope in Badong urban area,Three Gorges Reservoir area[J]. Resources Environment & Engineering,2023,37(3):292 − 300. (in Chinese with English abstract)] CHEN Guojin. Stability investigation and risk control of bank slope in Badong urban area, Three Gorges Reservoir area[J]. Resources Environment & Engineering, 2023, 37(3): 292 − 300. (in Chinese with English abstract)
[3] 张曦,陈丽霞,徐勇,等. 两种斜坡单元划分方法对滑坡灾害易发性评价的对比研究[J]. 安全与环境工程,2018,25(1):12 − 17. [ZHANG Xi,CHEN Lixia,XU Yong,et al. Comparison of two methods for slope unit division in landslide susceptibility evaluation[J]. Safety and Environmental Engineering,2018,25(1):12 − 17. (in Chinese with English abstract)] ZHANG Xi, CHEN Lixia, XU Yong, et al. Comparison of two methods for slope unit division in landslide susceptibility evaluation[J]. Safety and Environmental Engineering, 2018, 25(1): 12 − 17. (in Chinese with English abstract)
[4] 曾斌,吕权儒,寇磊,等. 基于Logistic回归和随机森林的清江流域长阳库岸段堆积层滑坡易发性评价[J]. 中国地质灾害与防治学报,2023,34(4):105 − 113. [ZENG Bin,LYU Quanru,KOU Lei,et al. Susceptibility assessment of colluvium landslides along the Changyang section of Qingjiang River using Logistic regression and random forest methods[J]. The Chinese Journal of Geological Hazard and Control,2023,34(4):105 − 113. (in Chinese with English abstract)] ZENG Bin, LYU Quanru, KOU Lei, et al. Susceptibility assessment of colluvium landslides along the Changyang section of Qingjiang River using Logistic regression and random forest methods[J]. The Chinese Journal of Geological Hazard and Control, 2023, 34(4): 105 − 113. (in Chinese with English abstract)
[5] 王凯,张少杰,韦方强. 斜坡单元提取方法研究进展和展望[J]. 长江科学院院报,2020,37(6):85 − 93. [WANG Kai,ZHANG Shaojie,WEI Fangqiang. Slope unit extraction methods:advances and prospects[J]. Journal of Yangtze River Scientific Research Institute,2020,37(6):85 − 93. (in Chinese with English abstract)] DOI: 10.11988/ckyyb.20190210 WANG Kai, ZHANG Shaojie, WEI Fangqiang. Slope unit extraction methods: advances and prospects[J]. Journal of Yangtze River Scientific Research Institute, 2020, 37(6): 85 − 93. (in Chinese with English abstract) DOI: 10.11988/ckyyb.20190210
[6] 刘军旗,刘强,刘千慧,等. 大数据时代地质灾害数据管理及应用模式探讨[J]. 地质科技通报,2021,40(6):276 − 282. [LIU Junqi,LIU Qiang,LIU Qianhui,et al. Discussion of geological hazard data management and application model in big data era[J]. Bulletin of Geological Science and Technology,2021,40(6):276 − 282. (in Chinese with English abstract)] LIU Junqi, LIU Qiang, LIU Qianhui, et al. Discussion of geological hazard data management and application model in big data era[J]. Bulletin of Geological Science and Technology, 2021, 40(6): 276 − 282. (in Chinese with English abstract)
[7] 刘宝生,陈刚,程刚建. 江苏南京地质灾害风险评价[J]. 中国地质灾害与防治学报,2023,34(4):97 − 104. [LIU Baosheng,CHEN Gang,CHENG Gangjian. Risk assessment of geological disasters in Nanjing,Jiangsu Province[J]. The Chinese Journal of Geological Hazard and Control,2023,34(4):97 − 104. (in Chinese with English abstract)] LIU Baosheng, CHEN Gang, CHENG Gangjian. Risk assessment of geological disasters in Nanjing, Jiangsu Province[J]. The Chinese Journal of Geological Hazard and Control, 2023, 34(4): 97 − 104. (in Chinese with English abstract)
[8] 王俊德,杜晓阳,黄天浩,等. 河南省嵩县地质灾害风险评价[J]. 中国地质灾害与防治学报,2023,34(4):86 − 96. [WANG Junde,DU Xiaoyang,HUANG Tianhao,et al. Risk assessment of geological hazards in Song County,Henan Province[J]. The Chinese Journal of Geological Hazard and Control,2023,34(4):86 − 96. (in Chinese with English abstract)] WANG Junde, DU Xiaoyang, HUANG Tianhao, et al. Risk assessment of geological hazards in Song County, Henan Province[J]. The Chinese Journal of Geological Hazard and Control, 2023, 34(4): 86 − 96. (in Chinese with English abstract)
[9] 牟家琦,庄建琦,王世宝,等. 基于深度神经网络模型的雅安市滑坡易发性评价[J]. 中国地质灾害与防治学报,2023,34(3):157 − 168. [MU Jiaqi,ZHUANG Jianqi,WANG Shibao,et al. Evaluation of landslide susceptibility in Ya’an City based on depth neural network model[J]. The Chinese Journal of Geological Hazard and Control,2023,34(3):157 − 168. (in Chinese with English abstract)] MU Jiaqi, ZHUANG Jianqi, WANG Shibao, et al. Evaluation of landslide susceptibility in Ya’an City based on depth neural network model[J]. The Chinese Journal of Geological Hazard and Control, 2023, 34(3): 157 − 168. (in Chinese with English abstract)
[10] 支泽民,刘峰贵,周强,等. 基于流域单元的地质灾害易发性评价——以西藏昌都市为例[J]. 中国地质灾害与防治学报,2023,34(1):139 − 150. [ZHI Zemin,LIU Fenggui,ZHOU Qiang,et al. Evaluation of geological hazards susceptibility based on watershed units:A case study of the Changdu City,Tibet[J]. The Chinese Journal of Geological Hazard and Control,2023,34(1):139 − 150. (in Chinese with English abstract)] ZHI Zemin, LIU Fenggui, ZHOU Qiang, et al. Evaluation of geological hazards susceptibility based on watershed units: A case study of the Changdu City, Tibet[J]. The Chinese Journal of Geological Hazard and Control, 2023, 34(1): 139 − 150. (in Chinese with English abstract)
[11] 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. [GUO Fei,WANG Xiujuan,CHEN Xi,et al. Comparative analyses on susceptibility of cutting slope landslides in southern Jiangxi using different models[J]. The Chinese Journal of Geological Hazard and Control,2022,33(6):125 − 133. (in Chinese with English abstract)] GUO Fei, WANG Xiujuan, CHEN Xi, et al. Comparative analyses on susceptibility of cutting slope landslides in southern Jiangxi using different models[J]. The Chinese Journal of Geological Hazard and Control, 2022, 33(6): 125 − 133. (in Chinese with English abstract)
[12] 黄发明,殷坤龙,蒋水华,等. 基于聚类分析和支持向量机的滑坡易发性评价[J]. 岩石力学与工程学报,2018,37(1):156 − 167. [HUANG Faming,YIN Kunlong,JIANG Shuihua,et al. Landslide susceptibility assessment based on clustering analysis and support vector machine[J]. Chinese Journal of Rock Mechanics and Engineering,2018,37(1):156 − 167. (in Chinese with English abstract)] HUANG Faming, YIN Kunlong, JIANG Shuihua, et al. Landslide susceptibility assessment based on clustering analysis and support vector machine[J]. Chinese Journal of Rock Mechanics and Engineering, 2018, 37(1): 156 − 167. (in Chinese with English abstract)
[13] WANG Jiaqi,SUN Pengfei,CHEN Leilei,et al. Recent advances of deep learning in geological hazard forecasting[J]. Computer Modeling in Engineering & Sciences,2023,137(2):1381 − 1418.
[14] 夏辉,殷坤龙,梁鑫,等. 基于SVM-ANN模型的滑坡易发性评价——以三峡库区巫山县为例[J]. 中国地质灾害与防治学报,2018,29(5):13 − 19. [XIA Hui,YIN Kunlong,LIANG Xin,et al. Landslide susceptibility assessment based on SVM-ANN Models:a case stualy for Wushan County in the Three Gorges Reservoir[J]. The Chinese Journal of Geological Hazard and Control,2018,29(5):13 − 19. (in Chinese with English abstract)] XIA Hui, YIN Kunlong, LIANG Xin, et al. Landslide susceptibility assessment based on SVM-ANN Models: a case stualy for Wushan County in the Three Gorges Reservoir[J]. The Chinese Journal of Geological Hazard and Control, 2018, 29(5): 13 − 19. (in Chinese with English abstract)
[15] 何清,李宁,罗文娟,等. 大数据下的机器学习算法综述[J]. 模式识别与人工智能,2014,27(4):327 − 336. [HE Qing,LI Ning,LUO Wenjuan,et al. A survey of machine learning algorithms for big data[J]. Pattern Recognition and Artificial Intelligence,2014,27(4):327 − 336. (in Chinese with English abstract)] DOI: 10.3969/j.issn.1003-6059.2014.04.007 HE Qing, LI Ning, LUO Wenjuan, et al. A survey of machine learning algorithms for big data[J]. Pattern Recognition and Artificial Intelligence, 2014, 27(4): 327 − 336. (in Chinese with English abstract) DOI: 10.3969/j.issn.1003-6059.2014.04.007
[16] YOUSSEF K,SHAO K,MOON S,et al. Landslide susceptibility modeling by interpretable neural network[J]. Communications Earth and Environment,2023,4(1):162. DOI: 10.1038/s43247-023-00806-5
[17] NACHAPPA T,GHORBANZADEH O,GHOLAMNIA K,et al. Multi-hazard exposure mapping using machine learning for the state of Salzburg,Austria[J]. Remote Sensing,2020,12(17):2757. DOI: 10.3390/rs12172757
[18] 马晨曦. 后三峡时期库区城市人居环境建设评价研究——以巴东、秭归为例[D]. 重庆:重庆大学,2019. [MA Chenxi. Evaluation of urban human settlement environment construction in the reservoir area in the post-three gorges period[D]. Chongqing:Chongqing University,2019. (in Chinese with English abstract)] MA Chenxi. Evaluation of urban human settlement environment construction in the reservoir area in the post-three gorges period[D]. Chongqing: Chongqing University, 2019. (in Chinese with English abstract)
[19] 左可顺. 高切坡数据集成与高效管理研究——以三峡湖北库区为例[D]. 中国地质大学(武汉),2023. [ZUO Keshun. Research on Data Integration and Efficient Management of High Cutting Slope--Taking Three Gorges Hubei Reservoir Area as an Example[D]. China University of Geosciences (Wuhan),2023. (in Chinese)] ZUO Keshun. Research on Data Integration and Efficient Management of High Cutting Slope--Taking Three Gorges Hubei Reservoir Area as an Example[D]. China University of Geosciences (Wuhan), 2023. (in Chinese)
[20] 石菊松,张永双,董诚,等. 基于GIS技术的巴东新城区滑坡灾害危险性区划[J]. 地球学报,2005,26(3):275 − 282. [SHI Jusong,ZHANG Yongshuang,DONG Cheng,et al. GIS-based landslide hazard zonation of the new Badong County site[J]. Acta Geosicientia Sinica,2005,26(3):275 − 282. (in Chinese with English abstract)] DOI: 10.3321/j.issn:1006-3021.2005.03.014 SHI Jusong, ZHANG Yongshuang, DONG Cheng, et al. GIS-based landslide hazard zonation of the new Badong County site[J]. Acta Geosicientia Sinica, 2005, 26(3): 275 − 282. (in Chinese with English abstract) DOI: 10.3321/j.issn:1006-3021.2005.03.014
[21] 王瑛,林齐根,史培军. 中国地质灾害伤亡事件的空间格局及影响因素[J]. 地理学报,2017,72(5):906 − 917. [WANG Ying,LIN Qigen,SHI Peijun. Spatial pattern and influencing factors of casualty events caused by landslides[J]. Acta Geographica Sinica,2017,72(5):906 − 917. (in Chinese with English abstract)] DOI: 10.11821/dlxb201705011 WANG Ying, LIN Qigen, SHI Peijun. Spatial pattern and influencing factors of casualty events caused by landslides[J]. Acta Geographica Sinica, 2017, 72(5): 906 − 917. (in Chinese with English abstract) DOI: 10.11821/dlxb201705011
[22] 吴润泽,程温鸣,刘军旗,等. 三峡库区地质灾害防治信息系统及预警指挥系统数据管理模式探讨[J]. 中国地质灾害与防治学报,2018,29(5):102 − 107. [WU Runze,CHENG Wenming,LIU Junqi,et al. Discussion on the data management mode of geologic disaster prevention and control information system and early warning command system in the Three Gorges Reservoir Area[J]. The Chinese Journal of Geological Hazard and Control,2018,29(5):102 − 107. (in Chinese with English abstract)] WU Runze, CHENG Wenming, LIU Junqi, et al. Discussion on the data management mode of geologic disaster prevention and control information system and early warning command system in the Three Gorges Reservoir Area[J]. The Chinese Journal of Geological Hazard and Control, 2018, 29(5): 102 − 107. (in Chinese with English abstract)
[23] LIU Junqi,HUANG Xuebin,WU Chonglong,et al. From the area to the point-study on the key technology of 3D geological hazard modeling in Three Gorges Reservoir area[J]. Journal of Earth Science,2012,23(2):199 − 206. DOI: 10.1007/s12583-012-0246-5
[24] 刘越凡,付萧,朱庆,等. 顾及地貌形态特征的精细斜坡单元高效分区划分[J]. 测绘科学,2023,48(4):211 − 220. [LIU Yuefan,FU Xiao,ZHU Qing,et al. An efficient and fine slope units division method with consideration of the regional geomorphological characteristics[J]. Science of Surveying and Mapping,2023,48(4):211 − 220. (in Chinese with English abstract)] LIU Yuefan, FU Xiao, ZHU Qing, et al. An efficient and fine slope units division method with consideration of the regional geomorphological characteristics[J]. Science of Surveying and Mapping, 2023, 48(4): 211 − 220. (in Chinese with English abstract)
[25] HUANG Faming,CAO Zhongshan,JIANG Shuihua,et al. Landslide susceptibility prediction based on a semi-supervised multiple-layer perceptron model[J]. Landslides,2020,17(12):2919 − 2930. DOI: 10.1007/s10346-020-01473-9
[26] 刘丽娜,许冲,徐锡伟,等. GIS支持下基于AHP方法的2013年芦山地震区滑坡危险性评价[J]. 灾害学,2014,29(4):183 − 191. [LIU Lina,XU Chong,XU Xiwei,et al. GIS-based landslide hazard evaluation using AHP method in the 2013 Lushan earthquake region[J]. Journal of Catastrophology,2014,29(4):183 − 191. (in Chinese with English abstract)] DOI: 10.3969/j.issn.1000-811X.2014.04.034 LIU Lina, XU Chong, XU Xiwei, et al. GIS-based landslide hazard evaluation using AHP method in the 2013 Lushan earthquake region[J]. Journal of Catastrophology, 2014, 29(4): 183 − 191. (in Chinese with English abstract) DOI: 10.3969/j.issn.1000-811X.2014.04.034
[27] 李松林,许强,汤明高,等. 三峡库区滑坡空间发育规律及其关键影响因子[J]. 地球科学,2020,45(1):341 − 354. [LI Songlin,XU Qiang,TANG Minggao,et al. Study on spatial distribution and key influencing factors of landslides in Three Gorges Reservoir area[J]. Earth Science,2020,45(1):341 − 354. (in Chinese with English abstract)] LI Songlin, XU Qiang, TANG Minggao, et al. Study on spatial distribution and key influencing factors of landslides in Three Gorges Reservoir area[J]. Earth Science, 2020, 45(1): 341 − 354. (in Chinese with English abstract)
[28] 石菊松,徐瑞春,石玲,等. 基于RS和GIS技术的清江隔河岩库区滑坡易发性评价与制图[J]. 地学前缘,2007,14(6):119 − 128. [SHI Jusong,XU Ruichun,SHI Ling,et al. ETM+ imagery and GIS-based landslide susceptibility mapping for the regional area of Geheyan Reservoir on the Qingjiang River,Hubei Province,China[J]. Earth Science Frontiers,2007,14(6):119 − 128. (in Chinese with English abstract)] DOI: 10.3321/j.issn:1005-2321.2007.06.015 SHI Jusong, XU Ruichun, SHI Ling, et al. ETM+ imagery and GIS-based landslide susceptibility mapping for the regional area of Geheyan Reservoir on the Qingjiang River, Hubei Province, China[J]. Earth Science Frontiers, 2007, 14(6): 119 − 128. (in Chinese with English abstract) DOI: 10.3321/j.issn:1005-2321.2007.06.015
[29] WANG Jinge,XIANG Wei,LU Ning. Landsliding triggered by reservoir operation:a general conceptual model with a case study at Three Gorges Reservoir[J]. Acta Geotechnica,2014,9(5):771 − 788. DOI: 10.1007/s11440-014-0315-2
[30] 郭惠娟,唐南奇,林金宝. 基于GIS的仙游县土地利用与滑坡灾害敏感性分析[J]. 福建农林大学学报(自然科学版),2010,39(4):417 − 420. [GUO Huijuan,TANG Nanqi,LIN Jinbao. Sensibility analysis of land-use and landslide hazard based on GIS in Xianyou County[J]. Journal of Fujian Agriculture and Forestry University (Natural Science Edition),2010,39(4):417 − 420. (in Chinese with English abstract)] GUO Huijuan, TANG Nanqi, LIN Jinbao. Sensibility analysis of land-use and landslide hazard based on GIS in Xianyou County[J]. Journal of Fujian Agriculture and Forestry University (Natural Science Edition), 2010, 39(4): 417 − 420. (in Chinese with English abstract)
[31] 章昱,王磊,伏永朋,等. 基于斜坡单元与信息量法的丹江口库区典型流域地质灾害易发性评价[J]. 华南地质,2023,39(3):512 − 522. [ZHANG Yu,WANG Lei,FU Yongpeng,et al. Evaluation of geological disaster susceptibility in typical watershed of Danjiangkou Reservoir area based on slope unit and information method[J]. South China Geology,2023,39(3):512 − 522. (in Chinese with English abstract)] DOI: 10.3969/j.issn.2097-0013.2023.03.010 ZHANG Yu, WANG Lei, FU Yongpeng, et al. Evaluation of geological disaster susceptibility in typical watershed of Danjiangkou Reservoir area based on slope unit and information method[J]. South China Geology, 2023, 39(3): 512 − 522. (in Chinese with English abstract) DOI: 10.3969/j.issn.2097-0013.2023.03.010
[32] 刘亮,杨洋. 基于网格单元和随机森林的滑坡易发性评价[J]. 石化技术,2023,30(6):180 − 182. [LIU Liang,YANG Yang. Landslide susceptibility evaluation based on grid cells and random forest[J]. Petrochemical Industry Technology,2023,30(6):180 − 182. (in Chinese with English abstract)] DOI: 10.3969/j.issn.1006-0235.2023.06.060 LIU Liang, YANG Yang. Landslide susceptibility evaluation based on grid cells and random forest[J]. Petrochemical Industry Technology, 2023, 30(6): 180 − 182. (in Chinese with English abstract) DOI: 10.3969/j.issn.1006-0235.2023.06.060
[33] 陈刚,郝社锋,蒋波等. 基于机载LiDAR技术植被茂密区小型滑坡识别与评价[J/OL]. 自然资源遥感,2023,1 − 10. [CHEN Gang,HE Shefeng,JIANG Bo,et al. Identification and evaluation of small landslides in densely vegetated areas based on airborne LiDAR technology [J/OL]. Remote Sensing of Natural Resources,2023,1 − 10. (in Chinese with English abstract)] CHEN Gang, HE Shefeng, JIANG Bo, et al. Identification and evaluation of small landslides in densely vegetated areas based on airborne LiDAR technology [J/OL]. Remote Sensing of Natural Resources, 2023, 1 − 10. (in Chinese with English abstract)
[34] 张玺国,周雄冬,徐梦珍,等. 西藏地质灾害易发性及对水能开发适宜度影响[J]. 地理学报,2022,77(7):1603 − 1614. [ZHANG Xiguo,ZHOU Xiongdong,XU Mengzhen,et al. Distribution of hydropower development suitability in Tibet in the face of geological hazard susceptibility[J]. Acta Geographica Sinica,2022,77(7):1603 − 1614. (in Chinese with English abstract)] DOI: 10.11821/dlxb202207003 ZHANG Xiguo, ZHOU Xiongdong, XU Mengzhen, et al. Distribution of hydropower development suitability in Tibet in the face of geological hazard susceptibility[J]. Acta Geographica Sinica, 2022, 77(7): 1603 − 1614. (in Chinese with English abstract) DOI: 10.11821/dlxb202207003
[35] 邹浩,贾琳,郑路路等. 基于覆盖土层厚度识别的区域斜坡降雨入渗稳定性定量评价[J/OL]. 地球科学,2023,1 − 14. [Zhou Hao,Jia Lin,Zheng Lulu,et al. Quantitative evaluation of rainfall infiltration stability on regional slopes based on overburden soil thickness identification[J/OL]. Earth Science,2023,1 − 14. (in Chinese with English abstract)] Zhou Hao, Jia Lin, Zheng Lulu, et al. Quantitative evaluation of rainfall infiltration stability on regional slopes based on overburden soil thickness identification[J/OL]. Earth Science, 2023, 1 − 14. (in Chinese with English abstract)
[36] 丁世飞,齐丙娟,谭红艳. 支持向量机理论与算法研究综述[J]. 电子科技大学学报,2011,40(1):2 − 10. [DING Shifei,QI Bingjuan,TAN Hongyan. An overview on theory and algorithm of support vector machines[J]. Journal of University of Electronic Science and Technology of China,2011,40(1):2 − 10. (in Chinese with English abstract)] DING Shifei, QI Bingjuan, TAN Hongyan. An overview on theory and algorithm of support vector machines[J]. Journal of University of Electronic Science and Technology of China, 2011, 40(1): 2 − 10. (in Chinese with English abstract)
[37] WANG Sen,LING Sixiang,WU Xiyong,et al. Key predisposing factors and susceptibility assessment of landslides along the Yunnan–Tibet traffic corridor,Tibetan Plateau:comparison with the LR,RF,NB,and MLP techniques[J]. Frontiers in Earth Science,2023,10:1100363. DOI: 10.3389/feart.2022.1100363
[38] YU Haiwei,PEI Wenjie,ZHANG Jingyi,et al. Landslide susceptibility mapping and driving mechanisms in a vulnerable region based on multiple machine learning models[J]. Remote Sensing,2023,15(7):1886. DOI: 10.3390/rs15071886
[39] RAGHU S,SRIRAAM N. Optimal configuration of multilayer perceptron neural network classifier for recognition of intracranial epileptic seizures[J]. Expert Systems with Applications,2017,89(C):205 − 221.
[40] TASER P Y. Application of bagging and boosting approaches using decision tree-based algorithms in diabetes risk prediction[C]//The 7th International Management Information Systems Conference. Basel Switzerland:MDPI,2021,89(C):205 − 221.
[41] 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. [DOU Jie,XIANG Zilin,XU Qiang,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)] DOU Jie, XIANG Zilin, XU Qiang, 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)
[42] 孙滨,祝传兵,康晓波,等. 基于信息量模型的云南东川泥石流易发性评价[J]. 中国地质灾害与防治学报,2022,33(5):119 − 127. [SUN Bin,ZHU Chuanbing,KANG Xiaobo,et al. Susceptibility assessment of debris flows based on information model in Dongchuan,Yunnan Province[J]. The Chinese Journal of Geological Hazard and Control,2022,33(5):119 − 127. (in Chinese with English abstract)] SUN Bin, ZHU Chuanbing, KANG Xiaobo, et al. Susceptibility assessment of debris flows based on information model in Dongchuan, Yunnan Province[J]. The Chinese Journal of Geological Hazard and Control, 2022, 33(5): 119 − 127. (in Chinese with English abstract)
[43] NAMAN KAUR,HIMANSHU. Logistic regression:A basic approach[J]. Information and Communication Technol- ogy for Competitive Strategies,2022,623:481 − 488.
[44] 吴润泽,胡旭东,梅红波,等. 基于随机森林的滑坡空间易发性评价——以三峡库区湖北段为例[J]. 地球科学,2021,46(1):321 − 330. [WU Runze,HU Xudong,MEI Hongbo,et al. Spatial susceptibility assessment of landslides based on random forest:A case study from Hubei section in the Three Gorges Reservoir area[J]. Earth Science,2021,46(1):321 − 330. (in Chinese with English abstract)] WU Runze, HU Xudong, MEI Hongbo, et al. Spatial susceptibility assessment of landslides based on random forest: A case study from Hubei section in the Three Gorges Reservoir area[J]. Earth Science, 2021, 46(1): 321 − 330. (in Chinese with English abstract)