A Robust Deep-Learning Model for Landslide Susceptibility Mapping: A Case Study of Kurdistan Province, Iran

被引:43
作者
Ghasemian, Bahareh [1 ]
Shahabi, Himan [1 ]
Shirzadi, Ataollah [2 ]
Al-Ansari, Nadhir [3 ]
Jaafari, Abolfazl [4 ]
Kress, Victoria R. [5 ]
Geertsema, Marten [6 ]
Renoud, Somayeh [7 ]
Ahmad, Anuar [8 ]
机构
[1] Univ Kurdistan, Fac Nat Resources, Dept Geomorphol, Sanandaj 6617715175, Iran
[2] Univ Kurdistan, Fac Nat Resources, Dept Rangeland & Watershed Management, Sanandaj 6617715175, Iran
[3] Lulea Univ Technol, Civil Environm & Nat Resources Engn, S-97187 Lulea, Sweden
[4] Agr Res Educ & Extens Org AREEO, Res Inst Forests & Rangelands, Tehran 1496813111, Iran
[5] Univ Northern British Columbia, Dept Ecosyst Sci & Management, 3333 Univ Way, Prince George, BC V2N 4Z9, Canada
[6] Minist Forests Lands Nat Resource Operat & Rural, 499 George St, Prince George, BC V2L 1R5, Canada
[7] Univ Tehran, Coll Farabi, Dept Engn, Data Min Lab, Tehran 1417935840, Iran
[8] Univ Teknol Malaysia UTM, Fac Built Environm & Surveying, Dept Geoinformat, Johor Baharu 81310, Malaysia
关键词
landslide susceptibility; extreme learning machine; deep belief network; genetic algorithm; GIS; Iran; RAINFALL-INDUCED LANDSLIDES; SUPPORT VECTOR MACHINES; DECISION TREE; LOGISTIC-REGRESSION; SPATIAL PREDICTION; FREQUENCY RATIO; RANDOM FOREST; GIS; HAZARD; ISLAND;
D O I
10.3390/s22041573
中图分类号
O65 [分析化学];
学科分类号
070302 ; 081704 ;
摘要
We mapped landslide susceptibility in Kamyaran city of Kurdistan Province, Iran, using a robust deep-learning (DP) model based on a combination of extreme learning machine (ELM), deep belief network (DBN), back propagation (BP), and genetic algorithm (GA). A total of 118 landslide locations were recorded and divided in the training and testing datasets. We selected 25 conditioning factors, and of these, we specified the most important ones by an information gain ratio (IGR) technique. We assessed the performance of the DP model using statistical measures including sensitivity, specificity, accuracy, F1-measure, and area under-the-receiver operating characteristic curve (AUC). Three benchmark algorithms, i.e., support vector machine (SVM), REPTree, and NBTree, were used to check the applicability of the proposed model. The results by IGR concluded that of the 25 conditioning factors, only 16 factors were important for our modeling procedure, and of these, distance to road, road density, lithology and land use were the four most significant factors. Results based on the testing dataset revealed that the DP model had the highest accuracy (0.926) of the compared algorithms, followed by NBTree (0.917), REPTree (0.903), and SVM (0.894). The landslide susceptibility maps prepared from the DP model with AUC = 0.870 performed the best. We consider the DP model a suitable tool for landslide susceptibility mapping.
引用
收藏
页数:28
相关论文
共 133 条
[1]   A comparative study of support vector machine and logistic model tree classifiers for shallow landslide susceptibility modeling [J].
Abedini, Mousa ;
Ghasemian, Bahareh ;
Shirzadi, Ataollah ;
Dieu Tien Bui .
ENVIRONMENTAL EARTH SCIENCES, 2019, 78 (18)
[2]   Comparison of GIS-based landslide susceptibility models using frequency ratio, logistic regression, and artificial neural network in a tertiary region of Ambon, Indonesia [J].
Aditian, Aril ;
Kubota, Tetsuya ;
Shinohara, Yoshinori .
GEOMORPHOLOGY, 2018, 318 :101-111
[3]   A comparison study on the quantitative statistical methods for spatial prediction of shallow landslides (case study: Yozidar-Degaga Route in Kurdistan Province, Iran) [J].
Asadi, Mitra ;
Goli Mokhtari, Leila ;
Shirzadi, Ataollah ;
Shahabi, Himan ;
Bahrami, Shahram .
ENVIRONMENTAL EARTH SCIENCES, 2022, 81 (02)
[4]  
Avali Viji R, 2014, AMIA Annu Symp Proc, V2014, P266
[5]   Landslides in Sado Island of Japan: Part II. GIS-based susceptibility mapping with comparisons of results from two methods and verifications [J].
Ayalew, L ;
Yamagishi, H ;
Marui, H ;
Kanno, T .
ENGINEERING GEOLOGY, 2005, 81 (04) :432-445
[6]   Changes in land cover and shallow landslide activity:: A case study in the Spanish Pyrenees [J].
Beguería, S .
GEOMORPHOLOGY, 2006, 74 (1-4) :196-206
[7]   Confluence effects in rivers: Interactions of basin scale, network geometry, and disturbance regimes [J].
Benda, L ;
Andras, K ;
Miller, D ;
Bigelow, P .
WATER RESOURCES RESEARCH, 2004, 40 (05) :W054021-W0540215
[8]   Stochastic forcing of sediment supply to channel networks from landsliding and debris flow [J].
Benda, L ;
Dunne, T .
WATER RESOURCES RESEARCH, 1997, 33 (12) :2849-2863
[9]   Characterising performance of environmental models [J].
Bennett, Neil D. ;
Croke, Barry F. W. ;
Guariso, Giorgio ;
Guillaume, Joseph H. A. ;
Hamilton, Serena H. ;
Jakeman, Anthony J. ;
Marsili-Libelli, Stefano ;
Newham, Lachlan T. H. ;
Norton, John P. ;
Perrin, Charles ;
Pierce, Suzanne A. ;
Robson, Barbara ;
Seppelt, Ralf ;
Voinov, Alexey A. ;
Fath, Brian D. ;
Andreassian, Vazken .
ENVIRONMENTAL MODELLING & SOFTWARE, 2013, 40 :1-20
[10]   Rainfall event-based landslide susceptibility zonation mapping [J].
Bhandary, Netra Prakash ;
Dahal, Ranjan Kumar ;
Timilsina, Manita ;
Yatabe, Ryuichi .
NATURAL HAZARDS, 2013, 69 (01) :365-388