Phase Transitions in Heisenberg Magnets Induced by Uniaxial Anisotropy: Wang–Landau and Machine Learning Simulations

Abstract

The critical properties of a three-dimensional anisotropic Heisenberg model in an external field has been simulated for the first time using an approach combining the Wang–Landau algorithm with machine learning methods, DBSCAN clustering, and principal component analysis. The threshold value of the parameter Δ_c separating the regions of the decisive effect of uniaxial anisotropy has been determined.