Evaluation and comparison of LogitBoost Ensemble,Fisher’s Linear Discriminant Analysis,logistic regression and support vector machines methods for landslide susceptibility mapping     

Binh Thai Pham  Indra Prakash 《国际地球制图》2019,34(3):316-333

The purpose of this study was to investigate and compare the capabilities of four machine learning methods namely LogitBoost Ensemble (LBE), Fisher’s Linear Discriminate Analysis (FLDA), Logistic Regression (LR) and Support Vector Machines (SVM) to select the best method for landslide susceptibility mapping. A part of landslide prone area of Tehri Garhwal district of Uttarakhand state, India, was selected as a case study. Validation of models was carried out using statistical analysis, the chi square test and the Receiver Operating Characteristic (ROC) curve. Result analysis shows that the LBE has the highest prediction ability (AUC = 0.972) for landslide susceptibility mapping, followed by the SVM (0.945), the LR (0.873) and the FLDA (0.870), respectively. Therefore, the LBE is the best and a promising method in comparison to other three models for landslide susceptibility mapping.  相似文献   

Establishing an early warning system for flash floods in Hoang Su Phi District,Ha Giang Province,Vietnam     

Thach Ngoc Nguyen  Canh Xuan Pham  Huy Quoc Nguyen  Toan Ngo Bao Dang 《Singapore journal of tropical geography》2019,40(2):312-333

Hoang Su Phi is a mountainous district in Ha Giang province, Vietnam. When the rainy season arrives in these parts, flash floods (besides landslides), frequently occur in many areas, seriously affecting the socio‐economic condition of the district. Therefore, in this paper, we have developed an early warning system for flash floods, established based on a geomorphological and hydrological approach. The basic principle underlying this system is the fact that flash floods will often occur where there is high potential risk with sufficient rainfall. In the model, eight parameters of the basin were used to build a potential flash flood map. Using the spatial processing module in an open source software, early predictions from automatic weather stations were interpolated and processed online to produce a potential risk map. Depending on the threshold of precipitation values, the results determined locations where flash flood may occur at various flash flood risk indices (FFRI). The system may be applied to support provision of early flash flood warning up to 1?6 days in advance in the district, allowing the local government ample time to make appropriate decisions on the prevention and/or mitigation of damages caused by flash flood hazards.  相似文献