Learning crystal field parameters using convolutional neural networks
arXiv:2011.12911 · doi:10.21468/SciPostPhys.11.1.011
Abstract
We present a deep machine learning algorithm to extract crystal field (CF) Stevens parameters from thermodynamic data of rare-earth magnetic materials. The algorithm employs a two-dimensional convolutional neural network (CNN) that is trained on magnetization, magnetic susceptibility and specific heat data that is calculated theoretically within the single-ion approximation and further processed using a standard wavelet transformation. We apply the method to crystal fields of cubic, hexagonal and tetragonal symmetry and for both integer and half-integer total angular momentum values of the ground state multiplet. We evaluate its performance on both theoretically generated synthetic and previously published experimental data on CeAgSb, PrAgSb and PrMgCu, and find that it can reliably and accurately extract the CF parameters for all site symmetries and values of considered. This demonstrates that CNNs provide an unbiased approach to extracting CF parameters that avoids tedious multi-parameter fitting procedures.
arXiv comment19 pages, 9 figures
References in corpus (18)
- Machine learning and the physical sciences
- A high-bias, low-variance introduction to Machine Learning for physicists
- Topological Kondo Insulators
- Dynamical Mean-Field Theory within the Full-Potential Methods: Electronic structure of Ce-115 materials
- A Theory of Anisotropic Semiconductor of Heavy Fermions
- Hastatic Order in URu2Si2
- Artificial Neural Network Approach to the Analytic Continuation Problem
- Unified Molecular Field Theory for Collinear and Noncollinear Heisenberg Antiferromagnets
- Crystal field states in a Quantum Spin Ice]{Crystal field states of Pr in the candidate quantum spin ice PrSnO
- Forward and Inverse Design of Kirigami via Supervised Autoencoder
- Preserved entropy and fragile magnetism
- Kondo-lattice ferromagnets and their peculiar order along the magnetically hard axis
- GTPack: A Mathematica group theory package for application in solid-state physics and photonics
- Can competition between crystal field and Kondo effect cause non-Fermi liquid like behavior?
- Evolution of Quantum Criticality in CeNi_{9-x}Cu_xGe_4
- Field-induced Ferrohastatic Order in Cubic Non-Kramers Doublet Systems
- Physical properties of single crystalline MgCu ( = Y, Ce-Nd, Gd-Dy, Yb) and the search for in-plane magnetic anisotropy in hexagonal systems
- Crystal fields and Kondo effect: specific heat for Cerium compounds