A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Assessing Landslide Susceptibility in Korea Using a Deep Neural Network
Rainfall is a key triggering factor for landslides. Most of landslides in Korea were triggered by heavy rainfall. In this study, we used a deep neural network (DNN) to assess landslide spatial probability at Mt. Hwangnyeong, Busan, Korea. The results was validated based on 26 landslides using a receiver operating characteristic (ROC) curves. The areas under the curve (AUC) of the success-rate curve and predicted-rate curve showed that the proposed model was successful in predicting the spatial probability of landslide at Mt. Hwangnyeong. In addition, the DNN model was compared to the infinite slope model and showed better performance than the infinite slope model. The performance of the DNN model at three different activation functions were also compared to select the optimum function. This result showed that the DNN model with ReLu function has the best accuracy. A classified landslide susceptibility (CLS) map was established from the landslide spatial probability map by the geometrical interval method. A statistical test was performed and indicated that the classified landslide susceptibility map had statistical significance.
Assessing Landslide Susceptibility in Korea Using a Deep Neural Network
Rainfall is a key triggering factor for landslides. Most of landslides in Korea were triggered by heavy rainfall. In this study, we used a deep neural network (DNN) to assess landslide spatial probability at Mt. Hwangnyeong, Busan, Korea. The results was validated based on 26 landslides using a receiver operating characteristic (ROC) curves. The areas under the curve (AUC) of the success-rate curve and predicted-rate curve showed that the proposed model was successful in predicting the spatial probability of landslide at Mt. Hwangnyeong. In addition, the DNN model was compared to the infinite slope model and showed better performance than the infinite slope model. The performance of the DNN model at three different activation functions were also compared to select the optimum function. This result showed that the DNN model with ReLu function has the best accuracy. A classified landslide susceptibility (CLS) map was established from the landslide spatial probability map by the geometrical interval method. A statistical test was performed and indicated that the classified landslide susceptibility map had statistical significance.
Assessing Landslide Susceptibility in Korea Using a Deep Neural Network
Lecture Notes in Civil Engineering
Reddy, J. N. (editor) / Wang, Chien Ming (editor) / Luong, Van Hai (editor) / Le, Anh Tuan (editor) / Nguyen, Ba-Quang-Vinh (author) / Do, Thanh-Hai (author) / Kim, Yun-Tae (author)
2022-09-21
12 pages
Article/Chapter (Book)
Electronic Resource
English
Deep neural network , Infinite slope model , Landslide susceptibility , Rainfall Energy , Sustainable Architecture/Green Buildings , Structural Materials , Geotechnical Engineering & Applied Earth Sciences , Building Construction and Design , Construction Management , Environmental Policy , Engineering
Landslide susceptibility mapping in Injae, Korea, using a decision tree
| Elsevier | 2010
Landslide susceptibility mapping in Injae, Korea, using a decision tree
| British Library Online Contents | 2010