Advanced AI & ML for Scientific Applications

This advanced course teaches how modern AI and ML methods are adapted for scientific data and discovery. You will learn how to handle small, noisy, and high-dimensional scientific datasets, build graph neural networks for molecules and networks, use physics-informed neural networks and surrogate models, apply generative models and transformers to scientific data, quantify uncertainty, and use active learning and Bayesian optimisation to guide experiments. The course also covers interpretability, validation, and reproducibility in research. Prior knowledge of Python, machine learning, and deep learning concepts is required.