Comparison of convolutional neural networks for landslide susceptibility mapping in Yanshan County, China

被引:388
作者
Wang, Yi [1 ]
Fang, Zhice [1 ]
Hong, Haoyuan [2 ,3 ,4 ]
机构
[1] China Univ Geosci, Inst Geophys & Geomat, Wuhan 430074, Hubei, Peoples R China
[2] Nanjing Normal Univ, Minist Educ, Key Lab Virtual Geog Environm, Nanjing 210023, Jiangsu, Peoples R China
[3] State Key Lab Cultivat Base Geog Environm Evolut, Nanjing 210023, Jiangsu, Peoples R China
[4] Jiangsu Ctr Collaborat Innovat Geog Informat Reso, Nanjing 210023, Jiangsu, Peoples R China
基金
中国国家自然科学基金;
关键词
Landslide susceptibility; Deep teaming; Convolutional neural network; Data presentation algorithm; Yanshan County; SUPPORT VECTOR MACHINE; SPATIAL PREDICTION; LOGISTIC-REGRESSION; ROTATION FOREST; HYBRID INTEGRATION; HIMALAYAN AREA; DECISION TREE; MODELS; MAPS; CLASSIFICATION;
D O I
10.1016/j.scitotenv.2019.02.263
中图分类号
X [环境科学、安全科学];
学科分类号
08 ; 0830 ;
摘要
Assessments of landslide disasters are becoming increasingly urgent. The aim of this study is to investigate a convolutional neural network (CNN) framework for landslide susceptibility mapping (LSM) in Yanshan County, China. The two primary contributions of this study arc summarized as follows. First. W the best of our knowledge, this report describes the first time that the CNN framework is used for LSM.Second, different data representation algorithms arc developed to construct three novel CNN architectures. In this work, sixteen influencing factors associated with landslide occurrence were considered and historical landslide locations were randomly divided into training (70% of the total) and validation (30%) sets. Validation of these CNNs was performed using different commonly used measures in comparison to several of the most popular machine learning and deep learning methods. The experimental results demonstrated that the proportions of highly susceptible zones in all of the CNN landslide susceptibility maps are highly similar and lower than 30%, which indicates that these CNNs are more practical for landslide prevention and management than conventional methods. Furthermore, the proposed CNN framework achieved higher or comparable prediction accuracy. Specifically, the proposed CNNs were 3.94%-7.45% and 0.079-0.151 higher than those of the optimized support vector machine (SVM) in terms of overall accuracy (OA) and Matthews correlation coefficient (MCC), respectively. (C) 2019 Elsevier B.V. All rights reserved.
引用
收藏
页码:975 / 993
页数:19
相关论文
共 71 条
[1]   An introduction to MCMC for machine learning [J].
Andrieu, C ;
de Freitas, N ;
Doucet, A ;
Jordan, MI .
MACHINE LEARNING, 2003, 50 (1-2) :5-43
[2]  
[Anonymous], 2004, COMBINING PATTERN CL
[3]  
[Anonymous], ENVIRON GEOL
[4]  
[Anonymous], 2003, BEST PRACTICES CONVO
[5]  
[Anonymous], 2017, GEOCARTO INT, DOI DOI 10.1080/10106049.2016.1140824
[6]  
[Anonymous], 2015, Nature, DOI [10.1038/nature14539, DOI 10.1038/NATURE14539]
[7]  
[Anonymous], 1984, NAT HAZARDS
[8]  
[Anonymous], MODELING EARTH SYSTE
[9]  
[Anonymous], ADV NEURAL INFORM PR, DOI DOI 10.1109/TPAMI.2016.2577031
[10]  
[Anonymous], ADV NEURAL INFORM PR