Data-Driven Discovery for the Design of Soft Matter Systems

Zachary Nicolau, University of Washington
Soft condensed matter systems exhibit fascinating and often nonconventional pattern formation underlying the function of many natural systems, as in the development of biological structure during embryogenesis and information processing in neuronal networks. We have only recently begun to explore the potential for the design and control of soft matter, but the inherent complexity poses a significant challenge for traditional, reductionist approaches. Recently, novel data-driven and machine-learning methods have emerged, promising to aid the discovery of effective and even parsimonious models directly from experimental and numerical data. These approaches may enable the design of transformative technological paradigms for computation, material engineering, and self-assembly, aiming to leverage rather than mitigate complexity. In this talk, I will discuss recent advances in automated model discovery and highlight their potential applications for design in soft matter systems.