Skip to article frontmatterSkip to article content

Calendar

WeekTopicDetailsPracticalAssignmentQuiz
1OrientationCourse arrangement
Overview of the course Open-source software
git, Github & Gitlab
Setup Python environment and use jupyter notebook
Git, Github & Gitlab
2CNY Week
3Computer and Computationmaterials informatics, computer, programming language, performance, high-performance computing (HPC)
open-sourced software
4Databasedata structure, database format, materials database, data generation, materials selection
visualization
data mining how to query database
data visualization
data mining
query data from materials project
simple data visualization and data mining (bulk modulus)
5Atomistic Structures Imolecules, graph, biochemistry, nanomaterials, crystalline materials, noncrystalline materials, ordering, prototypes, Wyckoff positions, structure map
symmetry, group theory, point groups, space groups, primitive cells, conventional cells, Brillouin zone, reciprocal space & path
6Atomistic Structures IIdefect
interface
grain boundary
use pymatgen, ase, and spglib to build models and symmetry analysis
use VESTA and ase-gui for structure visualization
11
7Models and Theories Imultiscale modelling, atomistic modelling, force field, codes
molecular dynamics and monte carlo
2
8Models and Theories IIquantum mechanics, electronic structure, density functional theory (DFT)
functionals, plane waves and pseudopotentials, codes
use ase to load force field and do some computation (equation of state, stability of diamond and graphite)
use ase to run molecular dynamics
9Optimizationintroduction, energy landscapes
local optimization: 1D, 2D & beyond optimization, performance
gradients, steepest descent, curvature
global optimization: genetic algorithm, basin hopping, stimulus annealing
no free lunch rule
3
10High-throughput Simulationthermodynamics and kinetics
finite temperature and entropy
convex hulls
phase diagram
AIRSS
Genetic Algorithm
2
11Machine Learning Iregression and fitting
neural networks
Decision tree & random Forest
classification
principal component analysis
clustering
use pytorch learn to run some simple classification
training, errors
4
12Machine Learning IIGraph neural network
Diffusion model
Pytorch for GNNs
13Machine Learning Potentialdescriptor: SOAP, ACE, GAP and etc.
errors
MACE
MACE examples35