Seminar Calendar, Information and Innovation Management
Learn the score
25 March 2026 | 2:00 p.m.— 3:00 p.m.
KK 1119, K. K. Leung Building, HKU
SPEAKER
Prof. Richard J. Samworth
Professor of Statistical Science
Director of the Statistical Laboratory
Centre for Mathematical Sciences | University of Cambridge
ABSTRACT
Score estimation has recently emerged as a key modern statistical challenge, due to its pivotal role in generative modelling via diffusion models. Moreover, it is an essential ingredient in a new approach to linear regression via convex $M$-estimation, where the corresponding error densities are log-concave. I will outline the antitonic score matching framework that underpins this latter application, and explain its advantages over ordinary least squares, for both estimation and inference (e.g. prediction intervals). Motivated by both problems, I will then present new results on the minimax risk of score estimation over classes of log-concave densities.
BIOGRAPHY
Richard Samworth obtained his PhD in Statistics from the University of Cambridge in 2004, and has remained in Cambridge since, becoming a full professor in 2013 and the Professor of Statistical Science in 2017. His main research interests are in nonparametric and high-dimensional statistics, as well as the statistical foundations of AI; he has developed methods and theory for shape-constrained inference, missing data, subgroup selection, deep learning, data perturbation techniques, changepoint estimation, variable selection and independence testing. Richard received the COPSS Presidents’ Award in 2018, was elected as a Fellow of the Royal Society in 2021 and was awarded the David Cox Medal for Statistics in 2025. He served as co-editor of the Annals of Statistics (2019-2021) and is currently IMS President-Elect.
