Shallow Landslide Susceptibility Models Based on Artificial Neural Networks Considering the Factor Selection Method and Various Non-Linear Activation Functions

Lee, Deukhwan; Kim, Yun-Tae; Lee, Seung-Rae

doi:10.3390/rs12071194

Cited by 52 publications

(28 citation statements)

References 52 publications

(76 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…These sites are both located outside of the Alpine region (Jura mountains and the foothills of the Alps) and on average, have gentler slopes (Baulmes 14 • and Hornbach 21 • ). Steeper slopes tend to be more susceptible to shallow landslides, which is in agreement with other studies that have found slope to be one of their top predictors (Budimir et al, 2015;Goetz et al, 2015;Tien Bui et al, 2016;Oh and Lee, 2017;Persichillo et al, 2017;Lombardo and Mai, 2018;Lee et al, 2020;Nhu et al, 2020b, a).…”

Section: Individual Site Modelssupporting

confidence: 92%

“…The explanatory variables selected for the statistical evaluation of the shallow landslide points are a combination of variables commonly found in landslide or shallow landslide susceptibility studies (Budimir et al, 2015;Chen et al, 2017;Cignetti et al, 2019;Kavzoglu et al, 2014;Lee et al, 2020;Meusburger and Alewell, 2009;Persichillo et al, 2017;Nhu et al, 2020b, a) and climate-related variables that may explain differences between the sites (e.g., strong precipitation events) from the CHELSA data set (Karger et al, 2017(Karger et al, , 2018. Variables related to land-cover and vegetation are not considered as we filter our study sites to contain only grassland areas.…”

Section: Explanatory Variablesmentioning

confidence: 99%

“…The aim of our study is to statistically evaluate shallow landslide occurrence for 10 different sites (between 16 and 54 km 2 ) across Switzerland. In the past, shallow landslide susceptibility studies have mainly focused on one or two study sites while often testing multiple modelling techniques (Gómez and Kavzoglu, 2005;Meusburger and Alewell, 2009;Vorpahl et al, 2012;Tien Bui et al, 2016;Oh and Lee, 2017;Lee et al, 2020;Nhu et al, 2020b) except for Persichillo et al (2017), who evaluated four sites in different catchments. For our shallow landslide inventory we map the eroded sites on aerial images (0.25 m resolution) using a U-Net deep learning approach (Ronneberger et al, 2015).…”

mentioning

confidence: 99%

See 2 more Smart Citations

Investigating Causal Factors of Shallow Landslides in Grassland Regions of Switzerland

Zweifel¹,

Samarin²,

Meusburger³

et al. 2021

Preprint

View full text Add to dashboard Cite

Abstract. Mountainous grassland slopes can be severely affected by soil erosion. To better understand the regional differences of soil erosion patterns, we determine the locations of shallow landslides across different sites and aim at identifying their triggering causal factors. Ten sites across Switzerland located in the Alps (8 sites), in foothill regions (1 site), and the Jura mountains (1 site) were selected for statistical evaluations. For the shallow landslide inventory, we used aerial images (0.25 m) with a deep learning approach (U-Net) to map the locations of eroded sites. We used logistic regression with a Group Lasso variable selection method to identify important explanatory variables for predicting the mapped shallow landslides. The set of variables consists of traditional susceptibility modelling factors and climate-related factors to represent local as well as cross-regional conditions. This set of explanatory variables (predictors) are used to develop individual site models (regional evaluation) as well as an all-in-one model (cross-regional evaluation) using all shallow landslide points simultaneously. While the local conditions of the ten sites lead to different variable selections, consistently slope and aspect were selected as the essential explanatory variables of shallow landslide susceptibility. Accuracy scores range between 70.2 and 79.8 % for individual site models. The all-in-one model confirms these findings by selecting slope, aspect as well as roughness as the most important explanatory variables (Accuracy = 72.3 %). Our finding suggest that traditional susceptibility variables describing geomorphological and geological conditions yield satisfactory results for all tested regions. However, for two sites with lower model accuracy, important processes may be under-represented with the available explanatory variables. The regression models for sites with an east-west oriented valley axis performed slightly better than models for north-south oriented valleys, which may be due to the influence of exposition related processes. Additionally, model performance is higher for Alpine sites, suggesting that core explanatory variables are understood for these areas.

show abstract

Section: Individual Site Modelssupporting

confidence: 92%

Section: Explanatory Variablesmentioning

confidence: 99%

mentioning

confidence: 99%

See 1 more Smart Citation

Investigating Causal Factors of Shallow Landslides in Grassland Regions of Switzerland

Zweifel¹,

Samarin²,

Meusburger³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…In addition, quantitative evaluation is the most widely used method in evaluating the susceptibility of geological hazards, such as information value model (Achour et al 2017;Liao et al 2019;Mokhtari and Abedian 2019;Sharma and Mahajan 2019;Chen et al 2020), mathematical statistics method (Bourenane et al 2016;Tien Bui et al 2016;Liao et al 2019;Sharma and Mahajan 2019;Chen et al 2020), certainty factor Zheng et al 2020), logistic regression (LR) (Wang et al 2018;Hu et al 2020), artificial neural network (ANN) (Xu et al 2015;Tien Bui et al 2016;Zhou, et al 2018;Sun et al 2019;Lee et al 2020;Wang et al 2020), decision tree (Zhang et al 2017;Hong et al 2018;Wang et al 2018), support vector machines (SVM) (Li et al 2015;Pham et al 2016;Tien Bui et al 2016;Wang et al 2018;Chen et al 2019;Mokhtari and Abedian 2019;Nguyen et al 2020;Yu and Gao 2020), and so on. These methods mainly use mathematical model to establish the quantitative relationship between geological hazards and evaluation factors, which can quantitatively describe the sensitivity of each evaluation factor in different intervals.…”

Section: Introductionmentioning

confidence: 99%

Optimization of random forest model for assessing and predicting geological hazards susceptibility in Lingyun County

Kong¹,

Wang²,

Ma³

et al. 2021

Preprint

View full text Add to dashboard Cite

The frequent occurrence of geological hazards will not only cause peoples' property loss and deterioration of living environments, but will also endanger peoples' lives. Therefore, rapid and accurate evaluation of geological hazards susceptibility can provide an important scientific basis for emergency rescue and disaster reduction and prevention. In this paper, ten effective variables including slope, aspect, curvature, normalized differential vegetation index, annual precipitation, strata lithology, tectonic complexity, residential density, road network density, and land use/land cover were selected as evaluation indexes. Meanwhile, random forest (RF) model is improved by the optimization of unbalanced geological hazards dataset, differentiation of continuous geological hazards evaluation factors, sample similarity calculation, and iterative method for finding optimal random characteristics by calculating out-of-bagger errors. The geological hazards susceptibility evaluation model based on optimized RF (OPRF) was established and used to assess the susceptibility level of geological hazards for Lingyun County. Then, receiver operating characteristics (ROC) curves and field investigation were performed to verify the efficiency for five models. Analysis and comparison of the results denoted that the model based on OPRF has the highest prediction accuracy of 93.4%, which is far better than the other four models. Furthermore, the evaluation results can provide reference for geological hazards prediction and prevention, and can also provide decision support for land use development and rational utilization of resources and environment in Lingyun County. Based on these results, the OPRF model could be extended to other regions with similar geological environment backgrounds for geological hazards susceptibility assessment and prediction.

show abstract

“…The Geographical Information System (GIS) and remote sensing methods have been used to assess landslide hazards. The landslide hazard maps have been produced using a variety of mathematical techniques, ranging from conventional statistic methods such as frequency ratio (Chen et al, 2020), statistical index and weights-of-evidence (Regmi et al, 2014), and logistic regression (Lombardo & Mai, 2018;Sun et al, 2018) to more recent advanced intelligence methods such as artificial neural network (ANN) (Shahri et al, 2019;Alkhasawneh et al, 2013;Alkhasawneh et al, 2014;Lee et al, 2020;Ortiz & Martínez-Graña, 2018). ANN is applied to many natural science applications such as speech recognition, human face recognition, classification of satellite images, and recognition of texture.…”

mentioning

confidence: 99%

Landslide Hazard Analysis Using a Multilayered Approach Based on Various Input Data Configurations

Huqqani¹,

Tay²,

Mohamad–Saleh³

2021

Geos. Ind.

View full text Add to dashboard Cite

Landslide is a natural disaster that occurs mostly in hill areas. Landslide hazard mapping is used to classify the prone areas to mitigate the risk of landslide hazards. This paper aims to compare spatial landslide prediction performance using an artificial neural network (ANN) model based on different data input configurations, different numbers of hidden neurons, and two types of normalization techniques on the data set of Penang Island, Malaysia. The data set involves twelve landslide influencing factors in which five factors are in continuous values, while the remaining seven are in categorical/discrete values. These factors are considered in three different configurations, i.e., original (OR), frequency ratio (FR), and mixed-type (MT) data, which act as an input to train the ANN model separately. A significant effect on the final output is the number of hidden neurons in the hidden layer. In addition, three data configurations are processed using two different normalization methods, i.e., mean-standard deviation (Mean-SD) and Min-Max. The landslide causative data often consist of correlated information caused by overlapping of input instances. Therefore, the principal component analysis (PCA) technique is used to eliminate the correlated information. The area under the receiver of characteristics (ROC) curve, i.e., AUC is also applied to verify the produced landslide hazard maps. The best result of AUC for both Mean-SD and Min-Max with PCA schemes are 96.72% and 96.38%, respectively. The results show that Mean-SD with PCA of MT data configuration yields the best validation accuracy, AUC, and lowest AIC at 100 number of hidden neurons. MT data configuration with the Mean-SD normalization and PCA scheme is more robust and stable in the MLP model's training for landslide prediction. Keywords: Landslide; ANN; Hidden Neurons; Normalization; PCA; ROC; Hazard map Copyright (c) 2021 Geosfera Indonesia and Department of Geography Education, University of Jember This work is licensed under a Creative Commons Attribution-Share A like 4.0 International License

show abstract

Shallow Landslide Susceptibility Models Based on Artificial Neural Networks Considering the Factor Selection Method and Various Non-Linear Activation Functions

Cited by 52 publications

References 52 publications

Investigating Causal Factors of Shallow Landslides in Grassland Regions of Switzerland

Investigating Causal Factors of Shallow Landslides in Grassland Regions of Switzerland

Optimization of random forest model for assessing and predicting geological hazards susceptibility in Lingyun County

Landslide Hazard Analysis Using a Multilayered Approach Based on Various Input Data Configurations

Contact Info

Product

Resources

About