日本金属学会 2020年春期(第166回)講演大会

Presentation information

Koubo Symposium

[S5] Element strategy for high performance permanent magnets 7 - Fundamental and basic research toward next-generation novel materials -(3)

Thu. Mar 19, 2020 9:00 AM - 1:50 PM Rm. H (W541,4th Flr., West Lecture Bldg.1)

Chair:Kanta Ono(KEK), Yoshihiro Gohda

11:55 AM - 12:10 PM

[S5.36] Bayesian optimization for searching magnet compounds and its application to data obtained by first-principles calculation

*Fukazawa Taro1,3, Harashima Yosuke1,2,3, Miyake Takashi1,3 (1. CD-FMat AIST, 2. iMaSS Nagoya Univ., 3. ESICMM NIMS)

Keywords:ThMn12型化合物、第一原理電子状態計算、機械学習、ベイズ最適化、永久磁石化合物探索

我々は機械学習と第一原理計算を組み合わせた磁石化合物の探索手法の開発を行っている。本発表ではThMn12型磁石化合物(R,Z)(Fe,Co,Ti)12 (R=Y, Nd, Sm; Z = Zr,Dy) の組成最適化を例にその探索効率の高さを示す。

Abstract password authentication.
Password is required to view the abstract. Please enter a password to authenticate.

Password