Automated discovery of characteristic features of phase transitions in many-body localization

We identify a new “order parameter” for the disorder-driven many-body localization transition by leveraging machine learning. Contrary to previous studies, our method is almost entirely unsupervised. A game theoretic process between neural networks defines an adversarial setup with conflicting objec...

Ausführliche Beschreibung

Bibliographische Detailangaben

Veröffentlicht in:	Physical review. B 2019-03, Vol.99 (10), p.104106, Article 104106
Hauptverfasser:	Huembeli, Patrick, Dauphin, Alexandre, Wittek, Peter, Gogolin, Christian
Format:	Artikel
Sprache:	eng
Schlagworte:	Artificial intelligence Automation Game theory Localization Machine learning Many body interactions Mapping Neural networks Order parameters Parameter identification Phase diagrams Phase transitions
Online-Zugang:	Volltext
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!