基于改进两步法采样策略和卷积神经网络的崩塌易发性评价  

Collapse susceptibility evaluation based on an improved two-step sampling strategy and a convolutional neural network

作　　者：邓日朗 张庆华[2] 刘伟[2] 陈凌伟[2] 谭建辉 高泽茂 郑先昌[1] DENG Rilang;ZHANG Qinghua;LIU Wei;CHEN Lingwei;TAN Jianhui;GAO Zemao;ZHENG Xianchang(School of Civil Engineering,Guangzhou University,Guangzhou 510006,China;Guangzhou Urban Planning&Design Survey Research Institute,Guangzhou 510060,China)

机构地区：[1]广州大学土木工程学院,广州510006 [2]广州市城市规划勘测设计研究院,广州510060

出　　处：《地质科技通报》2024年第2期186-200,共15页Bulletin of Geological Science and Technology

基　　金：广州市城市规划勘测设计研究院咨询项目(2023岩28008B-合01)。

摘　　要：机器学习在崩塌滑坡泥石流地质灾害易发性分析评价领域已得到广泛的研究性应用,非灾害样本的选取是易发性建模过程中的关键问题,传统随机抽样和手工标注方法可能存在随机性和主观性。将土质崩塌易发性评价视为正例无标记(positive and unlabeled,简称PU)学习,提出了一种结合信息量(information value,简称IV)和间谍技术(Spy)的两步卷积神经网络(convolutional neural networks,简称CNN)框架(ISpy-CNN)。以广州市黄埔区崩塌编录和15类基础环境因子,通过信息量模型筛选出部分低信息量样本;采用间谍技术训练CNN模型,从低信息量样本中识别出具有高置信度的可靠负例划分为非崩塌样本;分别基于该学习框架、传统间谍技术和随机抽样,使用支持向量机(support vector machine,简称SVM)和随机森林(random forest,简称RF)对比验证。结果表明,ISpy-CNN框架在验证集上的准确率、F1值、敏感度和特异度较随机采样分别提升了6.82%,6.82%,6.82%,8.23%,较传统Spy技术分别提升了2.86%,2.89%,2.86%,2.31%;PU学习中第2步采用CNN模型的预测精度高于RF和SVM模型;与传统Spy技术相比,增加相同数量训练样本,ISpy-CNN框架筛选的样本集表现出较高的稳定性、预测精度和增长率。本研究提出的ISpy-CNN框架能更好地辅助选取高质量非灾害样本,且崩塌易发性分区结果更符合实际的崩塌空间分布。[Objective] Machine learning has been widely applied in the fields of collapse,landslide and debris flow susceptibility analysis.The selection of nonhazard samples is a key issue in landslide susceptibility analysis.Traditional random sampling and manual labelling methods may involve randomness and subjectivity.[Methods] In view of the potential randomness and representativeness of noncollapse samples,this paper considered soil collapse susceptibility evaluation a positive-unlabelled(PU) learning problem and proposes a two-step convolutional neural network framework(ISpy-CNN) that combines an information value model and the Spy technique.First,15 collapse-related factors were selected for modelling based on the geomorphological,geological,hydrological,and artificial environmental conditions of the study area.Low-information-value samples that were able to map the distribution structure of noncollapsing samples were screened by the information value model.Then,through the Spy technique and training the CNN model,negative samples with high confidence were identified from low-information-value samples that were classified as noncollapsed samples.Finally,based on the framework and traditional random sampling,we used support vector machine(SVM) and random forest(RF) models to compare and verify the reliability,prediction accuracy and data sensitivity of the proposed learning framework and other models.[Results]The results illustrate that the proposed ISpy-CNN method can improve the accuracy,F1 value,sensitivity and specificity on the validation set by 6.82%,6.82%,6.82%,8.23%,respectively compared to random sampling and 2.86%,2.89%,2.86%,2.31%,respectively compared to the traditional Spy technique.The prediction accuracy of step 2 in PU learning using the CNN model is higher than that of the RF and SVM models.The sample set screened by the ISpy-CNN framework exhibited greater stability,prediction accuracy and growth rate than those screened by the traditional Spy technique by adding the same number of training samples.

关 键 词：崩塌 易发性评价 PU学习 间谍技术 信息量 卷积神经网络 随机森林 支持向量机 

分 类 号：P642.22[天文地球—工程地质学]