Training the next generation of materials scientists in AI and data-driven research
AI for Materials (Spring 2026)
Course Numbers: EN.510.463 (Undergraduate) & EN.510.663 (Graduate)
This course introduces core concepts and modern techniques in artificial intelligence and data science, with emphasis on applications in materials science. Students will learn to apply cutting-edge AI methods to real materials discovery challenges.
Course Topics
Fundamentals
- Classification, regression, and clustering
- Random forests and decision trees
- Model evaluation and validation
- Feature engineering
Deep Learning
- Neural networks (NN)
- Convolutional neural networks (CNN)
- Graph neural networks (GNN)
- Transformer architectures (GPT models)
Materials Applications
- Property prediction from structure
- Inverse materials design
- Generative AI for materials
- Computer vision for microscopy
Hands-on Experience
Students will work with real-world materials datasets from JARVIS spanning:
- Tabular data: Formation energies, bandgaps, elastic constants
- Microscopy images: STEM and STM nanoscale imaging
- Crystal structures: Atomic coordinates and symmetries
- Scientific text: Literature mining and knowledge extraction
Final Project
Students will complete a capstone project applying AI methods to a materials discovery challenge of their choosing. Projects can involve property prediction, structure generation, image analysis, or text mining. Students will present their findings to the class and prepare a final report.
Course Resources
📺 Video Lectures
Comprehensive tutorials on materials informaticsWatch →
💻 Code Notebooks
Hands-on Google Colab tutorialsGitHub →
📊 Datasets
Materials database accessExplore →
Prerequisites
Recommended background:
- Basic programming experience (Python preferred)
- Introductory materials science or chemistry
- Linear algebra and calculus
- Statistics fundamentals helpful but not required