Materials Science
AI and machine learning for materials discovery, property prediction, and design.
Overview
Materials informatics applies ML to understand, predict, and design new materials. Key areas include:
- Property Prediction — Predicting material properties from structure
- Materials Discovery — Finding new materials with desired properties
- Neural Network Potentials — ML-based interatomic potentials
- Self-Driving Labs — Autonomous experimentation
Quick Start
| Goal | Resource |
|---|---|
| Learn materials ML | Resources — ML for Materials (ICL) |
| Hands-on tutorials | Tutorials — matminer, transformers |
| Explore tools | Tools — pymatgen, ASE |
| Find datasets | Datasets — Materials Project, AFLOW |
Communities
LeMaterial
Open-source collaborative project for materials research.
- Working Groups: Large Language Models, Generative Models, Benchmarks
- Format: Slack community + monthly meetings
Key Research Groups
- Aspuru-Guzik Group (Toronto) — Self-driving labs, materials
- Walsh Group (ICL) — ML for materials
- Jablonka/LlamaLab — LLMs for materials
- Sparks Group (Utah) — Materials informatics
Conferences
| Conference | Focus | Timing |
|---|---|---|
| MRS | Materials research | Spring/Fall |
| AI4Mat (NeurIPS) | ML for materials | December |