An Introduction to Diffusion and Flow Models - Lecture 3

Explore the fundamentals of diffusion and flow models in this third lecture from the Data Science: Probabilistic and Optimization Methods II program at the International Centre for Theoretical Sciences. Delivered by Dheeraj Nagaraj, this 65-minute academic lecture provides an introduction to these important generative modeling techniques that have become central to modern machine learning and data science. Learn about the mathematical foundations, theoretical principles, and practical applications of diffusion and flow models as part of a comprehensive program designed to illuminate the core probabilistic and optimization methods that underpin current successes and future breakthroughs in data science. The lecture is part of a broader workshop featuring bootcamp sessions on foundational topics, tutorials, and research talks covering innovative developments in reinforcement learning, generative modeling, causal inference, and advanced probability and optimization methods. Gain insights into how rigorous theory informs robust, adaptable systems while exploring the evolving theoretical landscape of data science alongside researchers, students, and practitioners from diverse backgrounds.