The effectiveness of data-driven landslide susceptibility mapping relies on data integrity and advanced geospatial analysis; however, selecting the most suitable method and identifying key regional factors remains a challenging task. To address this, this study assessed the performance of six machine learning models, including Convolutional Neural Networks (CNNs), Random Forest (RF), Categorical Boosting (CatBoost), their CNN-based hybrid models (CNN+RF and CNN+CatBoost), and a Stacking Ensemble (SE) combining CNN, RF, and CatBoost in mapping landslide susceptibility along the Karakoram Highway in northern Pakistan. Twelve geospatial factors were examined, categorized into Topography/Geomorphology, Land Cover/Vegetation, Geology, Hydrology, and Anthropogenic Influence. A detailed landslide inventory of 272 occurrences was compiled to train the models. The proposed stacking ensemble and hybrid models improve landslide susceptibility modeling, with the stacking ensemble achieving an AUC of 0.91. Hybrid modeling enhances accuracy, with CNN–RF boosting RF’s AUC from 0.85 to 0.89 and CNN–CatBoost increasing CatBoost’s AUC from 0.87 to 0.90. Chi-square (χ²) values (9.8–21.2) and p-values (<0.005) confirm statistical significance across models. This study identifies approximately 20.70% of the area as from high to very high risk, with the SE model excelling in detecting high-risk zones. Key factors influencing landslide susceptibility showed slight variations across the models, while multicollinearity among variables remained minimal. The proposed modeling approach reduces uncertainties, enhances prediction accuracy, and supports decision-makers in implementing effective landslide mitigation strategies.

Underground coal mining can induce substantial surface deformation and trigger associated geological hazards. However, the quantitative links between mining-induced deformation, stress redistribution, and the spatial pattern of landslide occurrence remain insufficiently understood, particularly in loess-covered mining regions. Taking the Hecaogou Coal Mine in the Zichang mining area of the Loess Plateau, China, as an example, this study uses a coupled framework that integrates multi-temporal Interferometric Synthetic Aperture Radar (InSAR) observations with three-dimensional FLAC3D numerical simulation. We found that surface deformation is primarily concentrated above and adjacent to the mined-out zones, with maximum cumulative deformation of −169.3 mm during March 2017 and December 2023. The stepwise excavation simulations reveal that vertical displacement and vertical compressive stress in the overlying strata increase continuously as mining advances, thereby promoting tensile–shear failure and surface subsidence, with the subsidence magnitude quantitatively increasing from 3.7 mm at 200 m depth to 162 mm at 1000 m depth. A strong agreement between InSAR-derived deformation and simulated deformation fields is demonstrated, confirming the reliability of the modeled deformation process. We also found that the landslide density exhibits a strong spatial correlation with surface deformation, with high-density zones clustering near the mined-out areas. These findings enhance our understanding of how underground coal mining reshapes surface stability and influences the spatial pattern of landslide occurrences in coal mining regions.