Abstract

Machine learning is becoming an integral part of condensed matter research, offering powerful new tools to model physical systems, uncover hidden patterns, and connect microscopic structure to macroscopic behavior. This mini-colloquium brings together experimental, theoretical, and computational researchers who apply machine learning across both soft and hard condensed matter.

Topics of the mini-colloquium include ML-based phase identification, ML-supported trajectory analysis, reverse engineering of micro-and nano structures and generative data sampling. A special emphasis will be placed on physics-aware dimensionality reduction and data compression for classical and quantum many-body systems. In some systems, such physically informed compression enables us to reveal mechanisms that remain inaccessible to traditional experimental or simulation approaches; in others, multi-scale compression is essential simply to model the relevant processes with sufficient accuracy.

By highlighting parallels between soft and hard condensed matter, the colloquium aims to foster exchange across communities and to identify general strategies for interpretable, transferable, and scalable modeling of materials.

Organizers

Name	Affiliation
Carina Karner	TU Wien
Markus Wallerberger	TU Wien
Emanuela Bianchi	TU Wien