I’m thrilled to announce that I’ll be presenting at the Online Seminar on Monte Carlo on October 22 from 11:30 AM – 12:30 PM EST! I’ll be discussing recent work on the No-U-Turn Sampler (NUTS), a very clever Markov chain Monte Carlo method widely used in probabilistic programming for sampling from probability distributions with continuously differentiable densities.
NUTS has been a game-changer with its self-tuning capability, and I’m looking forward to share our latest insights into why it work so well – and how we’re taking it even further. Confronting the complexities of NUTS has led us to several key innovations. Over the past year, we’ve made significant strides, including a deeper understanding of NUTS’ reversibility, the first mixing time guarantee, and solutions to longstanding challenges around locally adapting its step size.
From scaling in high dimensional spaces to overcoming mixing bottlenecks, this talk will explore the many layers of this pivotal MCMC method. What excites me most is the potential for the seminar’s discussions to inspire the next generation of theory and samplers for statistical computing. I’m truly grateful to the seminar organizers for this opportunity and look forward to engaging with the Monte Carlo community!
If you’d like to learn more or join the talk, check out the seminar series here: Online Seminar on Monte Carlo.