Profile approach for recognition of three-dimensional magnetic structures

Яковлев, И. А.; Sotnikov, O. M.; Мазуренко, В. В.

doi:10.1103/physrevb.99.024430

Cited by 14 publications

(7 citation statements)

References 31 publications

(37 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…17. It is also not limited by two-dimensional case, the profile procedure we used for preprocessing magnetization data gives reliable classification results in the static case for three-dimensional magnets with Dzyaloshinskii-Moriya interaction 9 . The generalization power of the network can be increased by adding more hidden neurons layers.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Recurrent Network Classifier for Ultrafast Skyrmion Dynamics

2019

Self Cite

View full text Add to dashboard Cite

By using the supervised learning we train a recurrent neural network to recognize and classify ultrafast magnetization processes realized in two-dimensional nanosystems with Dzyaloshinskii-Moriya interaction. Our focus is on the different types of skyrmion dynamics driven by ultrafast magnetic pulses. Each process is represented as a sequence of the sorted magnetization vectors inputted into the network. The trained network can perform an accurate classification of the skyrmionic processes at zero temperature in wide ranges of parameters that are the magnetic pulse width and damping factor. The network performance is also demonstrated on different types of unseen data including finite temperature processes. Our approach can be easily adapted for creating an autonomous control system on skyrmion dynamics for experiments or device operations. II. MODEL, METHOD AND NETWORKIn this work we use the following spin Hamiltonian to describe two-dimensional materials hosting skyrmionsHere J ij and D ij are the isotropic exchange interaction and Dzyaloshinskii-Moriya vector, respectively. S i is a arXiv:1907.01814v1 [cond-mat.str-el]

show abstract

Section: Discussionmentioning

confidence: 99%

“…Namely, such a sorting provides a very accurate separation of the magnetic phases in the static case as discussed in Ref. 9. It also gives us opportunity to use a moderate number of the hidden layer neurons that is 512 for process classification.…”

Section: Introductionmentioning

confidence: 98%

Recurrent Network Classifier for Ultrafast Skyrmion Dynamics

2019

Self Cite

View full text Add to dashboard Cite

show abstract

“…Related to this we expect an even higher fraction of noncollinear ground states, for which the need for calculation of the spin wave dispersions with the more general framework of linear spin wave theory for noncollinear magnets is desired [56,57]. Furthermore, recent developments of machine learning techniques for lattice models and spin Hamiltonians, as for instance a profile method for recognition of three-dimensional magnetic structures [71], determination of phase transition temperatures by means of self-organizing maps [72], and a support vector machines based method for multiclassification of phases [73], will be most useful for identification and classification of competing magnetic phases at finite temperature, and the corresponding phase transition temperatures.…”

Section: Discussionmentioning

confidence: 99%

Spin wave excitations of magnetic metalorganic materials

Hellsvik

Pérez

Geilhufe

et al. 2020

Phys. Rev. Materials

View full text Add to dashboard Cite

The Organic Materials Database (OMDB) is an open database hosting about 22,000 electronic band structures, density of states and other properties for stable and previously synthesized 3dimensional organic crystals. The web interface of the OMDB offers various search tools for the identification of novel functional materials such as band structure pattern matching and density of states similarity search. In this work the OMDB is extended to include magnetic excitation properties. For inelastic neutron scattering we focus on the dynamical structure factor S(Q, ω) which contains information on the excitation modes of the material. We introduce a new dataset containing atomic magnetic moments and Heisenberg exchange parameters for which we calculate the spin wave spectra and dynamic structure factor with linear spin wave theory and atomistic spin dynamics. We thus develop the materials informatics tools to identify novel functional organic and metalorganic magnets.

show abstract

“…A new paradigm to investigate phase transitions based on neural networks recently emerged. It has been shown that neural networks can classify spin configurations sampled from a Monte Carlo simulation and thus can function as order parameter identifiers 15,16 . The neural network is trained in a supervised fashion, i.e.…”

Section: Introductionmentioning

confidence: 99%

Finding hidden order in spin models with persistent homology

Olsthoorn,

Hellsvik,

Balatsky

2020

Preprint

View full text Add to dashboard Cite

Persistent homology (PH) is a relatively new field in applied mathematics that studies the components and shapes of discrete data. In this work, we demonstrate that PH can be used as a universal framework to identify phases in spin models, including hidden order such as spin nematic ordering and spin liquids. By converting a small number of spin configurations to barcodes we obtain a descriptive picture of configuration space. Using dimensionality reduction to reduce the barcode space to color space leads to a visualization of the phase diagram.

show abstract

Profile approach for recognition of three-dimensional magnetic structures

Cited by 14 publications

References 31 publications

Recurrent Network Classifier for Ultrafast Skyrmion Dynamics

Recurrent Network Classifier for Ultrafast Skyrmion Dynamics

Spin wave excitations of magnetic metalorganic materials

Finding hidden order in spin models with persistent homology

Contact Info

Product

Resources

About