An Automated and Explainable Machine Learning Model for Landslide Susceptibility Mapping

The complexity of model training and the difficulty in explaining prediction results greatly restrict the development of machine learning in the field of landslide susceptibility assessment． This study constructs a comprehensive explainable landslide susceptibility assessment model based on the SHAP-XGBoost algorithm, introducing “Explainable Artificial Intelligence (XAI)” and “Automated Machine Learning (AutoML)” into landslide susceptibility assessment research, achieving automated operation of complex model training, hyperparameter optimization, landslide susceptibility assessment mapping, and model explanation． The results of testing in the Fengjie County of the Three Gorges Reservoir Area at two scales, grid units, and slope units, demonstrate explainable automated landslide susceptibility assessment with high predictive accuracy． The model based on grid and slope units achieves AUC values of 0.875 and 0.873, respectively, with accuracy, precision, recall, and F1 scores all significantly higher than 0.5． The SHAP algorithm provides explanations for the model from both global and local perspectives, aiding in understanding the distribution characteristics of causative factors in model building and the occurrence patterns of landslide disasters. Additionally, the SHAP algorithm can explain the prediction results of individual evaluation units with high credibility． The research results provide important references for the study of automated machine learning and explainable models．