Published July 2018
| Version v1
Journal article
Simulation study of atom diffusion in amorphous materials with neural network potentials
Creators
- 1. National Inst. of Advanced Industrial Science and Technology (AIST), Tsukuba, Ibaraki (Japan)
- 2. Tokyo University, Graduate School of Engineering, Tokyo (Japan)
Description
We review our recent study of the machine learning potentials based on the Behler-Parrinello neural network scheme for atom diffusion in amorphous materials. We simplify the neural network potentials to apply to the dynamics of multi-component amorphous materials by taking the system characteristics into account. (author)
Additional details
Additional titles
- Original title (Japanese)
- ニューラルネットワークを用いたアモルファス物質中の原子拡散の研究
Publishing Information
- Journal Title
- Kotai Butsuri
- Journal Volume
- 53
- Journal Issue
- 7
- Series
- 雑誌名:固体物理
- Journal Page Range
- p. 389-399
- ISSN
- 0454-4544
INIS
- Country of Publication
- Japan
- Country of Input or Organization
- Japan
- INIS RN
- 49102696
- Subject category
- S75: CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND SUPERFLUIDITY;
- Descriptors DEI
- AMORPHOUS STATE; ATOMS; CHEMICAL BONDS; COMPUTERIZED SIMULATION; DEGREES OF FREEDOM; DIFFUSION; GAUSS FUNCTION; INFORMATION THEORY; INTERPOLATION; LEARNING; MONTE CARLO METHOD; NEURAL NETWORKS; REGRESSION ANALYSIS; SYMMETRY; TRANSFORMATIONS
- Descriptors DEC
- CALCULATION METHODS; FUNCTIONS; MATHEMATICAL SOLUTIONS; MATHEMATICS; NUMERICAL SOLUTION; SIMULATION; STATISTICS
Optional Information
- Notes
- 28 refs., 5 figs., 1 tab.