Ground subsidence along the riverbanks near the Yangtze River Delta has been accelerating due to human activities and other factors, seriously impacting various aspects of social development. Mapping susceptibility patterns and analyzing subsidence factors are crucial for effective management. This study focused on the Yangtze River riparian perimeter in Jiangsu Province, our study area. We assessed the importance of different factors using the random forest regression (RFR) model and the temporal convolution network (TCN). Additionally, we used GeoDetector to analyze the spatial relationship between sedimentation and potential drivers. Finally, we utilized the RFR and Maxent model to map susceptibility to sedimentation patterns in different risk zones. The study results show that the method effectively depicts the susceptibility to subsidence in each risk zone (44.18% and 32.56% for high and average risk zones, respectively). Anthropogenic factors mainly drive the subsidence-prone areas around the Yangtze River in Jiangsu. Groundwater extraction and soft soil thickness are the primary drivers of subsidence patterns in high-risk areas. In contrast, the main drivers of subsidence in other risk areas vary. These differences reflect the delayed effects of natural and anthropogenic factors on subsidence and the significant differences in how anthropogenic drivers affect the marginal effects of subsidence. Through susceptibility modeling and driver evaluation, this study reveals that establishing risk zones has improved our understanding of the impact of regional variations in environmental variables on subsidence. This understanding will facilitate the development of subsidence management strategies tailored to different regions.