17–21 Sept 2022
Faculty of Physics, Astronomy and Applied Computer Science; Jagiellonian University
Europe/Warsaw timezone
Registration and Abstract submission are CLOSED

First or second? An attempt to determine the order of a phase transition with machine learning methods.

18 Sept 2022, 16:54
3m
Faculty of Physics, Astronomy and Applied Computer Science; Jagiellonian University

Faculty of Physics, Astronomy and Applied Computer Science; Jagiellonian University

Łojasiewicza 11, 30-348 Kraków, Poland

Speaker

Mrs Monika Richter-Laskowska (Institute of Medical Technology and Equipment)

Description

The determination of the order of a phase transition can be quite a challenging task. In the thermodynamic limit the situation is clear: in the case of a first--order phase transition we observe discontinuity in the free energy at the critical point leading to the release of a latent heat. This phenomenon is not observable in the case of a continuous phase transition. Therefore, the results obtained with Monte Carlo computations, which simulate only the finite--size systems do not always provide an answer whether transition is of first or second--order. We attempt to overcome this issue by a careful analysis of the output obtained from the application of the learning by confusion scheme to the Potts, Blume--Capel and Falicov--Kimball models. We discover that in some cases the establishment of the order of a phase transition is plausible, but the results obtained in the case of a discontinuous phase transition strictly depend on the number of degrees of freedom and their character (quantum, classical).

Primary authors

Mrs Monika Richter-Laskowska (Institute of Medical Technology and Equipment) Dr Marcin Kurpas Prof. Maciej Maśka

Presentation materials

There are no materials yet.