Seismic-induced landslide disasters frequently result in the extensive burial of road and railway infrastructure, leading to the disruption of transportation networks and impeding critical lifeline projects. These events severely hinder the efficient execution of emergency rescue operations and pose a significant threat to public safety and property. Traditional manual disaster assessment and damage estimation methods are often insufficient to support the rapid assessment and repair decision-making processes required for transportation infrastructure in the aftermath of an earthquake. Consequently, an intelligent and rapid assessment and simulation methodology lever-aging big data analytics is of paramount importance. Herein, based on existing disaster cases, this study elucidates the deformation evolution law of post-seismic land-slides and constructs a dataset for the deformation associated with earthquake-induced landslides via numerical inversion. Thereafter, an integrated predictive model is proposed, which combines with ensemble empirical mode decomposition and deep learning algorithms. Employing dimensionality reduction techniques, this model effectively eliminates redundant features within the dataset, facilitating the swift prediction of post-seismic landslide displacements and the identification of disaster-affected areas. Through the execution of model training and validation procedures on the dataset, the efficacy of the integrated predictive model has been thoroughly assessed. The findings demonstrate that the integrated model exhibits superior performance in terms of both predictive accuracy and computational speed, offering valuable theoretical insights and technical support for the expedited assessment of post-seismic landslide disaster.