Agent-based models are used widely in the modeling of disease processes. They have features that make them suitable as a modeling platform for spatially heterogeneous, multi-scale, stochastic disease processes, such as tumor growth or conditions involving the immune system. Since they are not equation-based and can be computationally expensive to simulate, they pose special challenges as models underlying digital twins. This talk will address mathematical challenges related to the construction of digital twins using agent-based models: data assimilation, optimal control, and the construction of surrogate models.
Copyright. All Rights Reserved.