Deep learning of phase transitions with minimal examples

Abstract

Over the past several years, there have been many studies demonstrating the ability of deep neural networks to identify phase transitions in many physical systems, notably in classical statistical physics systems. One often finds that the prediction of deep learning methods trained on many ensembles below and above the critical temperature $T_{\mathrm{c}}$ behaves similarly to an order parameter, and this analogy has been successfully used to locate $T_{\mathrm{c}}$ and estimate universal critical exponents. In this work, we pay particular attention to the ability of a convolutional neural network to capture these critical parameters for the 2-$d$ Ising model when the network is trained on configurations at $T=0$ and $T=\infty$ only. We directly compare its output to the same network trained at multiple temperatures below and above $T_{\mathrm{c}}$ to gain understanding of how this extreme restriction of training data can impact a neural network's ability to classify phases. We find that the network trained on two temperatures is still able to identify $T_{\mathrm{c}}$ and $\nu$, while the extraction of $\gamma$ becomes more challenging.

8 pages, 3 figures

References in corpus (27)

• SciPy 1.0--Fundamental Algorithms for Scientific Computing in Python
• The order of the quantum chromodynamics transition predicted by the standard model of particle physics
• Machine learning phases of matter
• Learning phase transitions by confusion
• Chiral crossover in QCD at zero and non-zero chemical potentials
• Discovering Phase Transitions with Unsupervised Learning
• Unsupervised learning of phase transitions: from principal component analysis to variational autoencoders
• The QCD crossover at finite chemical potential from lattice simulations
• Discovering Phases, Phase Transitions and Crossovers through Unsupervised Machine Learning: A critical examination
• Identifying topological order through unsupervised machine learning
• Bayesian model averaging for analysis of lattice field theory results
• Polyakov loop in 2+1 flavor QCD from low to high temperatures
• Applications of neural networks to the studies of phase transitions of two-dimensional Potts models
• Thermodynamics and Feature Extraction by Machine Learning
• Improved information criteria for Bayesian model averaging in lattice field theory
• Machine learning determination of dynamical parameters: The Ising model case
• Deep Learning on the 2-Dimensional Ising Model to Extract the Crossover Region with a Variational Autoencoder
• Extending machine learning classification capabilities with histogram reweighting
• A Performance Study of the 2D Ising Model on GPUs
• A Neural Networks study of the phase transitions of Potts model
• Model averaging approaches to data subset selection
• Charm degrees of freedom in hot matter from lattice QCD
• Streamlined data analysis in Python
• Dependence of the static quark free energy on ${\mu }_B$ and the crossover temperature of $N_f=2+1$ QCD
• Examples of renormalization group transformations for image sets
• A simple framework for contrastive learning phases of matter
• A Very Effective and Simple Diffusion Reconstruction for the Diluted Ising Model

Cited by in corpus (2)

• Comments on the minimal training set for CNN: a case study of the frustrated $J_1$-$J_2$ Ising model on the square lattice
• Discovering quasiorder parameters in the Potts model: A bridge between machine learning and critical phenomena