Automatically (or interactively with a user) building digital twins from data

This project will consider ways to build digital twins from data automatically or with user interaction. One possible research topic here could be scaling from parameter tuning, where a user defines the general structure of the model and an algorithm fits the parameters to data, then the algorithm is given increasing power to decide on the general system dynamics. Another potential research topic could be to build a simulation model, use it to generate data, then automatically create digital twins with the data.