Machine learning for sampling high-dimensional probability distributions in lattice field theory

<!--HTML--><p>As machine learning algorithms continue to enable and accelerate physics calculations, the development of problem-specific physics-informed machine learning approaches is becoming more sophisticated, impactful, and important. I will describe recent advances in generative modelling emerging from the challenge of exact sampling from known probability distributions in the context of lattice quantum field theory calculations in particle and nuclear physics. I will discuss in particular flow-based generative models, outline the importance of guarantees of exactness and the incorporation of complex symmetries (e.g., gauge symmetry) into model architectures, and show how this can be achieved.</p><p><i>Phiala Shanahan is an Associate Professor of physics at MIT. Her research group works to understand nuclear physics from the Standard Model through first-principles calculations using tools including lattice field theory, effective field theories, and many-body methods. She received her PhD from the University of Adelaide, spent two years as a postdoctoral fellow at MIT and one year as a joint faculty member at the College of William &amp; Mary and senior staff scientist at the Thomas Jefferson National Accelerator facility, before moving to her current position.</i></p><p><strong>Coffee will be served at 10:30.</strong></p>