The data-driven approach has become significant in various scientific fields, such as climate modeling and weather forecasting, where a mechanistic description of physics and chemistry is either ...

Differential privacy (DP) has seen immense applications in learning on tabular, image, and sequential data where instance-level privacy is concerned. In learning on graphs, contrastingly, works on ...