Search

Originating from image recognition, methods of machine learning allow for effective feature extraction and dimensionality reduction in multidimensional datasets, thereby providing an extraordinary tool to deal with classical and quantum models in many-body physics. In this study, we employ a specific unsupervised machine learning technique -- self-organizing maps -- to create a low-dimensional representation of microscopic states, relevant for macroscopic phase identification and detecting phase transitions. We explore the properties of spin Hamiltonians of two archetype model system: a two-dimensional Heisenberg ferromagnet and a three-dimensional crystal, Fe in the body centered cubic structure. The method of self-organizing maps, that is known to conserve connectivity of the initial dataset, is compared to the cumulant method theory and is shown to be as accurate while being computationally more efficient in determining a phase transition temperature. We argue that the method proposed here can be applied to explore a broad class of second-order phase transition systems, not only magnetic systems but also, for example, order-disorder transitions in alloys., Comment: 5 pages, 2 figures, Supplemental Material

By means of first principles calculations, we have systematically investigated how the magnetodynamic properties Gilbert damping, magnetization and exchange stiffness are affected when permalloy (Py) (Fe$_{0.19}$Ni$_{0.81}$) is doped with 4d or 5d transition metal impurities. We find that the trends in the Gilbert damping can be understood from relatively few basic parameters such as the density of states at the Fermi level, the spin-orbit coupling and the impurity concentration. % The temperature dependence of the Gilbert damping is found to be very weak which we relate to the lack of intraband transitions in alloys. % Doping with $4d$ elements has no major impact on the studied Gilbert damping, apart from diluting the host. However, the $5d$ elements have a profound effect on the damping and allows it to be tuned over a large interval while maintaining the magnetization and exchange stiffness. % As regards spin stiffness, doping with early transition metals results in considerable softening, whereas late transition metals have a minor impact. % Our result agree well with earlier calculations where available. In comparison to experiments, the computed Gilbert damping appears slightly underestimated while the spin stiffness show good general agreement., Comment: 8 pages, 9 figures