Professor David Blackwell (1919 – 2010) was a distinguished American statistician and mathematician who made significant contributions to game theory, probability theory, information theory, and Bayesian statistics. He is one of the eponyms of the Rao–Blackwell theorem and became the first African-American elected member of the US National Academy of Sciences. As a Department, we wanted to mark his ground-breaking work and are delighted to hold this annual lecture named after him in October each year.
Previous Lectures
Date: Thursday 26th October 2023
Speaker: Professor Eric J Tchetgen Tchetgen
Title: Introducing the Forster-Warmuth Nonparametric Counterfactual Regression
Abstract: Series regression estimation is one of the most popular non-parametric regression techniques in practice. The most routinely used series estimator is based on ordinary least squares fitting, which is known to be minimax rate optimal in various settings, albeit under stringent restrictions on the basis functions. In this work, inspired by the recently developed Forster-Warmuth (FW) regression, we propose an alternative nonparametric series estimator that can attain minimax estimation rates under strictly weaker conditions imposed on the basis functions, than virtually all existing series estimators in the literature. Another contribution of this work generalizes the FW-regression to a so-called counterfactual regression problem, in which the response variable of interest may not be directly observed (hence, the name ``counterfactual'') on all sampled units. Although counterfactual regression is not entirely a new area of inquiry, we propose the first-ever systematic study of this challenging problem from a unified pseudo-outcome perspective. In fact, we provide what appears to be the first generic and constructive approach for generating the pseudo-outcome (to substitute for the unobserved response) which leads to the estimation of the counterfactual regression curve of interest with small bias, namely bias of second order. Several applications are used to illustrate the resulting FW counterfactual regression including a large class of nonparametric regression problems in missing data and causal inference literature, for which we establish conditions for minimax rate optimality. This is joint work with Yachong Yang and Arun Kuchibhotla.
Date: Friday 21st October 2022
Speaker: Professor Sir Peter Donnelly
Title: A path to personalised disease prevention: using genomics to predict risk for common diseases
Abstract: It has long been known that genetics is a major risk factor for all the common chronic human diseases, such as heart disease, diabetes, osteoporosis, and the mental health and auto-immune disorders, and for the common cancers, such as breast, prostate, and bowel cancer. For many of these diseases, or in some cases for important subsets of individuals, it is the single most important predictor of disease risk. We now know, from 20 years of human genetics studies, that this risk is polygenic: for any given disease, it is due to the cumulative effects of a very large number of different genetic variants in our DNA, each of individually small impact. Polygenic risk scores (PRSs) provide a single measure which quantifies the overall impact of these variants. For many diseases the underlying genetic studies are now large enough to provide good information about which variants matter for a particular disease, and their effect sizes, and in parallel we now have large prospective population cohorts, such as UK Biobank, in which to assess and validate predictive power.
Date: October 2021 (Online)
Speaker: John Jumper, Senior Research Scientist, DeepMind
Date: Tuesday 6th October 2020 (Online)
Speaker: Jason Forrest, Director of Interactive Data Visualization, McKinsey & Co, New York
Title: Exploring the Data Visualizations of W.E.B. Du Bois
Abstract: At the 1900 Paris Exposition, an all African-American team lead by scholar and activist W.E.B. Du Bois sought to challenge and recontextualize the perception of African-Americans at the dawn of the 20th-century. In less than 5 months, his team conducted sociological research and hand-made more than 60 large data visualization posters for a massive European audience which ultimately awarded Du Bois a gold medal for his efforts. While relatively obscure until recently, the ramification of his landmark work remains challenging and especially important in light of the Black Lives Matter movement.