Combining Data-Driven and Physics-Based Approaches to Predict, Understand, and Control Active Matter Dynamics

Watch a seminar presentation exploring the intersection of data-driven and physics-based approaches in understanding active matter dynamics. Professor Michael Hagan from Brandeis University delves into the challenges of modeling active materials, particularly microtubule-based active nematics. Learn about two complementary approaches: using sparse regression to discover optimal continuum models from spatiotemporal data, and applying deep learning techniques to forecast active nematics dynamics. Discover how these methods can reveal physical mechanisms, estimate experimental parameters, and potentially control active materials toward specific emergent behaviors. The presentation covers the complexities of measurement errors, reduced-dimensional representations, and the application of control theory in driving active materials toward particular behaviors.