Supervised learning magnetic skyrmion phases

We propose a simple and transparent machine learning approach for recognition and classification of complex non-collinear magnetic structures in two-dimensional materials. It is based on the implementation of the single-hidden-layer neural network that only relies on the z projections of the spins. In this setup one needs a limited set of magnetic configurations to distiguish ferromagnetic, skyrmion and spin spiral phases, as well as their different combinations. The network trained on the configurations for square-lattice Heisenberg model with Dzyaloshinskii-Moriya interaction can classify the magnetic structures obtained from Monte Carlo calculations for triangular lattice. Our approach is also easy to use for analysis of the numerous experimental data collected with spin-polarized scanning tunneling experiments.