Presentation Information

[25a-61C-5]DMI Energy Cost Analysis in the Formation Process of Skyrmions by Extended Landau Free Energy Model

〇(B)Yotaro Machida1, Michiki Taniwaki1, Alexandre Foggiatto1, Chiharu Mitsumata1, Ippei Obayashi2, Yasuaki Hiraoka3, Kyoko Ishizaka4, Masato Kotsugi1 (1.Tokyo Univ. of Sci., 2.Okayama Univ., 3.Kyoto Univ., 4.Univ. of Tokyo)

Keywords:

skyrmion,DMI,machine learning

The extended Landau free energy model was expanded to the formation process of skyrmions, and persistent homology analysis was applied to spin textures to extract features. In machine learning, we quantitatively linked the relationship between structural change and energy using principal component analysis with high explanatory power. The relationship between the structural change of the spin texture and the DMI energy density change was drawn in data space, and the obtained features were found to be robustly connected to magnetization and DMI.