An Integrated Approach of Machine Learning, Remote Sensing, and GIS Data for the Landslide Susceptibility Mapping

Landslides triggered in mountainous areas can have catastrophic consequences, threaten human life, and cause billions of dollars economic losses. Hence, it is imperative to map the susceptible landslides minimize their risk. Around Abbottabad, a large city northern Pakistan, number be found. This study aimed landslide susceptibility over these regions Pakistan by using three Machine Learning (ML) techniques, specifically Linear Regression (LiR), Logistic (LoR), Support Vector (SVM). Several influencing factors were used identify potential areas, including elevation, slope degree, aspect, general curvature, plan profile landcover classification system, Normalized Difference Water Index (NDWI), Vegetation (NDVI), soil, lithology, fault density, topographic roughness index, road density. The weights calculated ML techniques. weightage overlay tool adopted final output. According models, NDWI, slope, LCCS significantly impact occurrence. area under ROC curve (AUC) applied validate performance results show AUC value LiR (88%) better than SVM (86%) LoR (85%) models. models gives good accuracy, which reliable for results. study’s outcome provides baselines policymakers propose adequate protection mitigation measures against region, any other researcher adopt this methodology another having similar characteristics.