A Novel Heterogeneous Ensemble Framework Based on Machine Learning Models for Shallow Landslide Susceptibility Mapping

Abstract

Landslides are devastating natural disasters that seriously threaten human life and property. Landslide susceptibility mapping (LSM) plays a key role in landslide hazard management. Machine learning (ML) models are widely used in LSM but suffer from limitations such as overfitting and unreliable accuracy. To improve the classification performance of a single machine learning (ML) model, this study selects logistic regression (LR), support vector machine (SVM), random forest (RF), and gradient boosting decision tree (GBDT), and proposes a novel heterogeneous ensemble framework based on Bayesian optimization (BO), namely, stratified weighted averaging (SWA), to test its applicability in a typical landslide area in Yanbian Prefecture, China. Firstly, a dataset consisting of 1531 historical landslides was collected from field investigations and historical records, and a spatial database containing 16 predisposing factors was established. The dataset was divided into a training set and a test set in a ratio of 7:3. The results showed that SWA effectively improved the Accuracy, AUC, and robustness of the model compared to a single ML model. The SWA achieved the best classification results (Accuracy = 91.39% and AUC = 0.967). To verify the generalization ability of SWA, we selected published landslide datasets from Yanshan country and Yongxin country in China for testing. SWA also performed well, with an AUC of 0.871 and 0.860, respectively. As indicated by shapely values (SVs), Normalized Difference Vegetation Index (NDVI) is the factor that has the greatest impact on landslide occurrence. The landslide susceptibility maps obtained from this study will provide an effective reference program for land use planning and disaster prevention and mitigation projects in Yanbian Prefecture, China.

Improving landslide susceptibility prediction through ensemble Recursive Feature Elimination and meta-learning framework

“…By comparing the results of the present study to previous research conducted in various study regions with similar topographical and geological conditions, using either single or hybrid models, it can be inferred that there is a disparity in the obtained AUC values and accuracy. Compared with the other methods we found that Random Subspace and Logistic Model Tree (RSLMT) (AUC = 0.815 and Accuracy = 0.738) (Luo et al, 2019), Bayesian optimization (BO), namely, Stratified Weighted Averaging (SWA) (AUC = 0.967 and Accuracy = 0.914) (Haozhe et al, 2023), novel ensemble DL model, namely GL-ResNet (Residual Network) (AUC = 0.960 and Accuracy = 0.909) (Li et al, 2023) the hybrid model used in the current study produced high performances for Sub-Himalayan As a fundamental statistical model, Logistic Regression (LR) provides a baseline for the prediction of landslide susceptibility. It is effective in modeling binary outcomes (landslide occurrence or non-occurrence) based on a set of predictor variables.…”

Integrating Machine Learning Ensembles for Landslide Susceptibility Mapping in Northern Pakistan

“…Random forest (RF) is an extensively assumed technique in regression as well as the stratification of tasks, exploiting an amalgamation of various decision trees for the sake of predictive analysis [70,71]. For classifications, RF utilizes majority voting to calculate class.…”

Landslide Hazard Assessment Combined with InSAR Deformation: A Case Study in the Zagunao River Basin, Sichuan Province, Southwestern China

“…Various technologies and methods may be used to create a landslide inventory, including drone photogrammetry, high-resolution satellite optical imagery, laser scanning interpretation, and field surveys [18,19]. Landslide susceptibility reflects the probability of a landslide occurring [5,20,21]. Landslide hazard assessments, based on susceptibility results, consider triggering factors such as rainfall or earthquakes [12].…”