Digital twins are real-time digital replicas of physical systems that enable simulation, monitoring, and optimization. Unlike traditional models, they are dynamic (continuously updated), bi-directional (inform physical systems), and predictive (enable forecasting).

Technical challenges

To create a …

Context

We present a general framework to combine existing models in a robust manner. Our method is flexible, places few assumptions on the models aggregated, and performs well in our experiments.

Abstract

Whether deterministic or stochastic, models can be viewed as functions designed to …

Talk on Model Aggregation framework for combining predictions from diverse models