DPG Phi
Verhandlungen
Verhandlungen
DPG

Berlin 2024 – wissenschaftliches Programm

Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

DY: Fachverband Dynamik und Statistische Physik

DY 19: Machine Learning in Dynamics and Statistical Physics II (joint session DY/SOE)

DY 19.3: Vortrag

Dienstag, 19. März 2024, 10:00–10:15, BH-N 243

Coarse-graining non-equilibrium systems with machine learning: from conceptual challenges to new approaches — •Patrick Egenlauf1,2 and Miriam Klopotek21University of Stuttgart, Interchange Forum for Reflecting on Intelligent Systems, IRIS3D project, Stuttgart, Germany — 2University of Stuttgart, Stuttgart Center for Simulation Science, SimTech Cluster of Excellence EXC 2075, Stuttgart, Germany

Machine learning (ML) was previously shown to effectively coarse-grain configurations of many-body systems. We want to investigate ML applications to address the dynamic coarse-graining of non-equilibrium many-body systems. Our research aims to advance ML methods while avoiding conventional assumptions. The focus is on time-dependent datasets and their broader implications for understanding causality. We introduce innovative techniques by incorporating general theory, including the time-dependent generalized Langevin equation [1], for building and interpreting time-dependent learning techniques [2]. This provides a distinctive ML perspective that extends to various applications for dynamical systems beyond equilibrium states. This study offers new ways to improve our understanding and manipulation of complex non-equilibrium many-body dynamics using ML.

[1] Schilling, T. (2022). Coarse-grained modelling out of equilibrium. Physics Reports, 972, 1-45.
[2] Nakajima, K., and Fischer, I. (2021). Reservoir Computing. Springer Singapore.

Keywords: Machine learning; Coarse-graining; Non-equilibrium; Time evolution; Causality

100% | Mobil-Ansicht | English Version | Kontakt/Impressum/Datenschutz
DPG-Physik > DPG-Verhandlungen > 2024 > Berlin