Classification of the non-Hermitian phases in a high-dimensional lattice becomes challenging due to interplay of the band topology and non-Hermiticity. The significant increase in data dimensions and the number of categories has rendered traditional supervised learning and unsupervised manifold learning unsuccessful. Here, we propose self-attention assistant machine learning for clustering non-Hermitian phases in a two-dimensional lattice. By incorporating the self-attention mechanism, the model can effectively capture long-range dependencies and important patterns, resulting in a more compact and information-rich latent space. It can achieve Altland-Zirnbauer classification with the Bloch vector dataset and distinguish the phases of eigenstates’ localized behavior with the competition between non-Hermitian skin effect and topological localization. Our results provide a general method for characterizing non-Hermitian phases in a two-dimensional lattice via machine learning.

Close

PrevNext

View figure in article

• Phase transitions
• Topological insulators
• Topological phases of matter
• Non-Hermitian systems
• Topological materials
• Deep learning
• Machine learning
• Unsupervised learning