English

Interpretable and unsupervised phase classification

Disordered Systems and Neural Networks 2021-07-21 v1 Strongly Correlated Electrons Quantum Physics

Abstract

Fully automated classification methods that yield direct physical insights into phase diagrams are of current interest. Here, we demonstrate an unsupervised machine learning method for phase classification which is rendered interpretable via an analytical derivation of its optimal predictions and allows for an automated construction scheme for order parameters. Based on these findings, we propose and apply an alternative, physically-motivated, data-driven scheme which relies on the difference between mean input features. This mean-based method is computationally cheap and directly interpretable. As an example, we consider the physically rich ground-state phase diagram of the spinless Falicov-Kimball model.

Keywords

Cite

@article{arxiv.2010.04730,
  title  = {Interpretable and unsupervised phase classification},
  author = {Julian Arnold and Frank Schäfer and Martin Žonda and Axel U. J. Lode},
  journal= {arXiv preprint arXiv:2010.04730},
  year   = {2021}
}

Comments

6+12 pages, 3+7 figures

R2 v1 2026-06-23T19:13:08.833Z