Dataset in the paper " From Model Optimization to Risk Representation: A Deep Learning Framework for Spatiotemporal Prediction of Mining-Induced Surface Deformation"

Surface deformation induced by natural resource extraction poses severe threats to both the environment and infrastructure. Spatiotemporal prediction using Interferometric Synthetic Aperture Radar (InSAR) time-series observations is critical for disaster prevention and mitigation. However, under strongly non-stationary spatial conditions, the complexity of hyperparameter settings in deep learning models and their associated prediction errors constrain their reliable application in practical risk quantification and management. To address this challenge, we developed an end-to-end workflow that spans from model optimization to risk representation. Detailed information can be obtained from the paper.