Sejong University

About: Sejong University is a education organization based out in Seoul, South Korea. It is known for research contribution in the topics: Graphene & Computer science. The organization has 5498 authors who have published 15236 publications receiving 330762 citations.

A novel ensemble decision tree-based CHi-squared Automatic Interaction Detection (CHAID) and multivariate logistic regression models in landslide susceptibility mapping

Abstract: An ensemble algorithm of data mining decision tree (DT)-based CHi-squared Automatic Interaction Detection (CHAID) is widely used for prediction analysis in variety of applications. CHAID as a multivariate method has an automatic classification capacity to analyze large numbers of landslide conditioning factors. Moreover, it results two or more nodes for each independent variable, where every node contains numbers of presence or absence of landslides (dependent variable). Other DT methods such as Quick, Unbiased, Efficient Statistic Tree (QUEST) and Classification and Regression Trees (CRT) are not able to produce multi branches based tree. Thus, the main objective of this paper is to use CHAID method to perform the best classification fit for each conditioning factors, then, combined it with logistic regression (LR) to find the corresponding coefficients of best fitting function that assess the optimal terminal nodes. In the first step, a landslide inventory map with 296 landslide locations were extracted from various sources over the Pohang-Kyeong Joo catchment (South Korea). Then, the inventory was randomly split into two datasets, 70 % was used for training the models, and the remaining 30 % was used for validation purpose. Thirteen landslide conditioning factors were used for the susceptibility modeling. Then, CHAID was applied and revealed that some conditioning factors such as altitude, soil drain, soil texture and TWI, as terminal nodes and reflected the best classification fit. Then, a proposed ensemble technique was applied and the interpretations of the coefficients showed that the relationship between the decision tree branch nodes distance from drain, soil drain, and TWI, respectively, leads to better consequences assessment of landslides in the current study area. The validation results showed that both success and prediction rates, 75 and 79 %, respectively. This study proved the efficiency and reliability of ensemble DT and LR model in landslide susceptibility mapping.

Spatial prediction of landslide susceptibility using data mining-based kernel logistic regression, naive Bayes and RBFNetwork models for the Long County area (China)

Abstract: The main goal of this study is to assess and compare three advanced machine learning techniques, namely, kernel logistic regression (KLR), naive Bayes (NB), and radial basis function network (RBFNetwork) models for landslide susceptibility modeling in Long County, China. First, a total of 171 landslide locations were identified within the study area using historical reports, aerial photographs, and extensive field surveys. All the landslides were randomly separated into two parts with a ratio of 70/30 for training and validation purposes. Second, 12 landslide conditioning factors were prepared for landslide susceptibility modeling, including slope aspect, slope angle, plan curvature, profile curvature, elevation, distance to faults, distance to rivers, distance to roads, lithology, NDVI (normalized difference vegetation index), land use, and rainfall. Third, the correlations between the conditioning factors and the occurrence of landslides were analyzed using normalized frequency ratios. A multicollinearity analysis of the landslide conditioning factors was carried out using tolerances and variance inflation factor (VIF) methods. Feature selection was performed using the chi-squared statistic with a 10-fold cross-validation technique to assess the predictive capabilities of the landslide conditioning factors. Then, the landslide conditioning factors with null predictive ability were excluded in order to optimize the landslide models. Finally, the trained KLR, NB, and RBFNetwork models were used to construct landslide susceptibility maps. The receiver operating characteristics (ROC) curve, the area under the curve (AUC), and several statistical measures, such as accuracy (ACC), F-measure, mean absolute error (MAE), and root mean squared error (RMSE), were used for the assessment, validation, and comparison of the resulting models in order to choose the best model in this study. The validation results show that all three models exhibit reasonably good performance, and the KLR model exhibits the most stable and best performance. The KLR model, which has a success rate of 0.847 and a prediction rate of 0.749, is a promising technique for landslide susceptibility mapping. Given the outcomes of the study, all three models could be used efficiently for landslide susceptibility analysis.