Assessing the importance of conditioning factor selection in landslide susceptibility for the province of Belluno (region of Veneto, northeastern Italy)

Abstract. In the domain of landslide risk science, susceptibility mapping (LSM) is very important, as it helps spatially identify potential landslide-prone regions. This study used a statistical ensemble model (frequency ratio and evidence belief function) two machine learning (ML) models (random forest XGBoost; eXtreme Gradient Boosting) for LSM in province Belluno (region Veneto, northeastern Italy). The investigated importance conditioning factors predicting occurrences using mentioned models. this paper, we evaluated overall prediction capabilities ML algorithms. By trial-and-error method, eliminated least “important” features by common threshold 0.30 0.03 Conclusively, found that removing important does not impact accuracy all three Based on results our study, most commonly available features, example, topographic contributes to comparable after ones, namely aspect plan profile curvature, wetness index (TWI), roughness (TRI), normalized difference vegetation (NDVI) case TWI, position (TPI) confirms requirement factor maps can be assessed based physiography region.