Landslide Susceptibility Mapping Using Shallow Neural Networks Model at Refahiye District in Turkey
Abstract
Landslides represent a continuous hazard for population and infrastructure. Mapping the landslide susceptibility is an essential issue to avoid the landslides risks. The aim of this paper is to produce a high-accuracy model for landslide susceptibility mapping in Refahiye district in Turkey. The model employed shallow neural networks for landslide susceptibility mapping, while bivariate spearman correlation test was utilized to select the related factors to extract the appropriate data and reduce the computation time of training and mapping. 12 out of 21 spatial factors were selected as relevant factors using Spearman correlation test. Relevant factors are geology, distance from roads, distance from geological faults, distance from water streams, flow direction, aspect, hillshade, heat load index, slope/aspect transformation, site exposure index, compound topographic index, and elevation. The generated dataset was divided into training, validation, and testing datasets using 10-folds cross-validation method. The TrainIm was found to be the best training function with an overall accuracy of 86.3%. The developed NN model was tested using IRIS benchmark dataset and showed higher performance against the logistic regression algorithm. As a result, shallow neural networks method was successfully applied in landslide susceptibility mapping in this study and the method is recommended for future studies.
Keywords
GIS , Landslide susceptibility mapping , Shallow neural networks
Kaynakça
- Can, A., Dagdelenler, G., Ercanoglu, M., & Sonmez, H. (2019). Landslide susceptibility mapping at Ovacık-Karabük (Turkey) using different artificial neural network models: comparison of training algorithms. Bulletin of Engineering Geology and the Environment, 78(1), 89–102.
- Chae, B.-G., Park, H.-J., Catani, F., Simoni, A., & Berti, M. (2017). Landslide prediction, monitoring and early warning: a concise review of state-of-the-art. Geosciences Journal, 21(6), 1033–1070.
- McCormick, C. (2016, February 24). K-fold cross-validation, with Matlab code. Retrieved from https://chrisjmccormick.wordpress.com/2013/07/31/k-fold-cross-validation-with-matlab-code/#
- Dağ, S., & Bulut, F. (2012). Coğrafi Bilgi Sistemleri Tabanlı Heyelan Duyarlılık Haritalarının Hazırlanmasına Bir Örnek: Çayeli (Rize, KD Türkiye). Jeoloji Mühendisliği Dergisi, 36(1), 35–62.
- Felicísimo, Á., Cuartero, A., Remondo, J., & Quirós, E. (2013). Mapping landslide susceptibility with logistic regression, multiple adaptive regression splines, classification and regression trees, and maximum entropy methods: a comparative study. Landslides, 10(2), 175–189.
- Hamad, R., Balzter, H., & Kolo, K. (2018). Predicting Land Use/Land Cover Changes Using a CA-Markov Model under Two Different Scenarios. Sustainability, 10(10), 3421, doi: 10.3390/su10103421
- JAXA. (2019, April 9). About ALOS - Overview and Objectives. Retrieved from https://www.eorc.jaxa.jp/ALOS/en/about/ about_index.htm
- Kim, D. E., & Gofman, M. (2018, January). Comparison of shallow and deep neural networks for network intrusion detection. In 2018 IEEE 8th Annual Computing and Communication Workshop and Conference (CCWC), 2018. pp. 204–208. IEEE.
- Lee, D.-H., Kim, Y.-T., & Lee, S.-R. (2020). Shallow Landslide Susceptibility Models Based on Artificial Neural Networks Considering the Factor Selection Method and Various Non-Linear Activation Functions. Remote Sensing, 12(7), 1194. doi: 10.3390/rs12071194
- Liang, D., Tsai, C.-F., & Wu, H.-T. (2015). The effect of feature selection on financial distress prediction. Knowledge-Based Systems, 73, 289–297.
