AI Materials Discovery
Machine learning and AI have become essential skills for computational materials scientists in the AI era. Here are some resources that have been particularly helpful to me:
Machine learning interatomic potential (MLIP) related
Selective Online Courses
Materials Database
- Mateirals Project, open web-based access to computed information on known and predicted materials
- JARVIS, Joint Automated Repository for Various Integrated Simulations
- NOMAD, open source data management platform
- Materials Cloud, Built for seamless sharing of resources in computational materials science
- …
High-throughput computational workflow
Some papers that have been inspiring recently
-
A generative model for inorganic materials design
- key words: MatterGen, a diffusion-based generative model, (stable, unique, novel) S.U.N, inverse design
- Active learning of reactive Bayesian force fields applied to heterogeneous catalysis dynamics of H/Pt
- key words: Sparse Gaussian process
-
E(3)-equivariant graph neural networks for data-efficient and accurate interatomic potentials
- key words: NequIP, E(3)-equivariant graph neural networks
-
Systematic softening in universal machine learning interatomic potentials
- key words: CHGNet, softing effect, fine-tuning
- Predicting Adsorption Energies for Catalyst Screening with Transfer Learning Using Crystal Hamiltonian Graph Neural Network