Mathematical Technology for Agent-Based Digital Twins

Explore mathematical challenges in developing agent-based digital twins for disease modeling in this 40-minute conference talk by Reinhard Laubenbacher from the University of Florida, presented at IPAM's Mathematics of Cancer workshop. Discover how agent-based models serve as powerful platforms for modeling spatially heterogeneous, multi-scale, and stochastic disease processes, including tumor growth and immune system conditions. Learn about the unique computational and mathematical challenges these non-equation-based models present when used as foundations for digital twins. Examine three critical mathematical components essential for constructing effective agent-based digital twins: data assimilation techniques for integrating real-world observations, optimal control methods for therapeutic interventions, and surrogate model construction approaches to address computational complexity. Gain insights into how these mathematical technologies can overcome the inherent challenges of computationally expensive agent-based simulations while maintaining their ability to capture complex biological phenomena across multiple scales and spatial heterogeneity.