Overview:
Catalysis is entering the era of artificial intelligence. This hands-on course empowers researchers to apply state-of-the-art machine learning techniques — from classical regression to deep neural networks — for predicting catalyst performance, accelerating materials discovery, and optimizing chemical reactions. Whether you’re a chemist or engineer, this course will help you integrate ML into your workflow with confidence.
What You’ll Learn:
- Core ML & AI concepts tailored for catalysis research
- Data types in catalysis: structural, spectral, kinetic, and compositional
- Data visualization and preprocessing techniques
- Feature engineering and descriptor extraction
- Supervised learning: regression, classification
- Unsupervised learning: clustering, dimensionality reduction
- Deep learning: neural networks for property prediction
- Physics-informed ML models and catalyst simulations
- Model refinement: hyperparameter tuning & overfitting control
- Practical lab: build and validate models with real catalytic data
- Tools: Python, pandas, scikit-learn, TensorFlow/PyTorch, ChemCatML
Course Learning Outcomes:
- Apply ML models (including deep learning) to catalysis datasets
- Analyze and visualize complex material data
- Use physics-informed ML to enhance prediction interpretability
- Build robust models and tune performance effectively
- Employ open-source ML tools for real catalyst problems
- Understand the limitations, ethics, and future trends of AI in catalysis
Who Should Join:
- Graduate students & researchers in catalysis or materials science
- Engineers applying AI in R&D
- Chemists interested in digital tools
- Data scientists entering the energy/materials field
Why Join:
- From theory to code — made for experimental scientists
- Unique focus on catalytic applications
- Based on real-world datasets and active research topics
- Learn ML with context — not just algorithms
- Future-proof your skills in AI-powered science