Paper Number
AM14 

Session
AI and ML Based Rheological Characterization

Title
Machine learning active-nematic hydrodynamics

Presentation Date and Time
October 10, 2022 (Monday) 4:25

Track / Room
Track 6 / Mayfair

Authors (Click on name to view author profile)

1. Han, Ming (University of Chicago)
2. Colen, Jonathan (University of Chicago)
3. Zhang, Rui (Hong Kong University of Science and Technology)
4. Redford, Steven A. (University of Chicago)
5. Lemma, Linnea M. (Brandeis Universitu)
6. Dogic, Zvonimir (University of California Santa Barbara)
7. Gardel, Margaret (University of Chicago)
8. Vitelli, Vincenzo (University of Chicago)
9. de Pablo, Juan J. (The University of Chicago, Pritzker School of Molecular Engineering)

Author and Affiliation Lines (in printed abstract book)
Ming Han¹, Jonathan Colen¹, Rui Zhang², Steven A. Redford¹, Linnea M. Lemma³, Zvonimir Dogic⁴, Margaret Gardel¹, Vincenzo Vitelli¹ and Juan J. de Pablo^5
1University of Chicago, Chicago, IL 60616; ²Hong Kong University of Science and Technology, Hong Kong, Hong Kong; ³Brandeis Universitu, Waltham, MA; ⁴University of California Santa Barbara, Santa Barbara, CA; ⁵Pritzker School of Molecular Engineering, The University of Chicago, Chicago, IL 60637

Speaker / Presenter 
Han, Ming

Keywords
computational methods; active matter; AI based; ML based

Text of Abstract
Hydrodynamic theories effectively describe many-body systems out of equilibrium in terms of a few macroscopic parameters. However, such parameters are difficult to determine from microscopic information. Seldom is this challenge more apparent than in active matter, where the hydrodynamic parameters are in fact fields that encode the distribution of energy-injecting microscopic components. Here, we use active nematics to demonstrate that neural networks can map out the spatiotemporal variation of multiple hydrodynamic parameters and forecast the chaotic dynamics of these systems. We analyze biofilament/molecular-motor experiments with microtubule/kinesin and actin/myosin complexes as computer vision problems. Our algorithms can determine how activity and elastic moduli change as a function of space and time, as well as adenosine triphosphate (ATP) or motor concentration. The only input needed is the orientation of the biofilaments and not the coupled velocity field which is harder to access in experiments. We can also forecast the evolution of these chaotic many-body systems solely from image sequences of their past using a combination of autoencoders and recurrent neural networks with residual architecture. In realistic experimental setups for which the initial conditions are not perfectly known, our physics-inspired machine-learning algorithms can surpass deterministic simulations. Our study paves the way for artificial-intelligence characterization and control of coupled chaotic fields in diverse physical and biological systems, even in the absence of knowledge of the underlying dynamics.