Our upcoming event for the Statistics Colloquium Series is scheduled for Monday, September 22 from 12:00 – 1:00pm (ET) and will be an in-person presentation Science Center 316. Lunch will be provided to guests following the talk. This week's speaker will be Veronika Rockova of Econometrics and Statistics at the University of Chicago Booth School of Business.
AI-Powered Bayesian Inference
The advent of Generative Artificial Intelligence (GAI) has heralded an inflection point that changed how society thinks about knowledge acquisition. While GAI cannot be fully trusted for decision-making, it may still provide valuable information that can be integrated into a decision pipeline. Rather than seeing the lack of certitude and inherent randomness of GAI as a problem, we view it as an opportunity. Indeed, variable answers to given prompts can be leveraged to construct a prior distribution which reflects assuredness of AI predictions. This prior distribution may be combined with tailored datasets for a fully Bayesian analysis with an AI-driven prior. In this paper, we explore such a possibility within a non-parametric Bayesian framework. The basic idea consists of assigning a Dirichlet process prior distribution on the data-generating distribution with AI generative model as its baseline. Hyper-parameters of the prior can be tuned out-of-sample to assess the informativeness of the AI prior. Posterior simulation is achieved by computing a suitably randomized functional on an augmented data that consists of observed (labeled) data as well as fake data whose labels have been imputed using AI. This strategy can be parallelized and rapidly produces iid samples from the posterior by optimization as opposed to sampling from conditionals. Our method enables (predictive) inference and uncertainty quantification leveraging AI predictions in a coherent probabilistic manner.
Veronika Rockova is Bruce Lindsay Professor of Econometrics and Statistics in the Wallman Society of Fellows at the Booth School of Business at the University of Chicago. She joined Booth after completing her postdoctoral training in statistics at the Wharton School of the University of Pennsylvania. Rockova has been internationally recognized for her broad research contributions at the intersection of statistics and machine learning. Her primary focus has been on creating innovative decision-centric tools for extracting insights from extensive datasets. She specializes in Bayesian computation, variable selection, high-dimensional decision theory, and hierarchical modeling. Her current research agenda includes uncertainty quantification and inference for generative models underlying many modern artificial intelligence platforms.
Her research was acknowledged with the National Science Foundation CAREER Award in 2020, the COPSS Emerging Leader Award in 2023 and the COPSS President's Award in 2024. This award is given annually by the Committee of Presidents of Statistical Societies to the “best statistician worldwide under 40 years of age”. She received her award for “path-breaking contributions to theory and methodology at the intersection of Bayesian and frequentist Statistics ... for exemplary service to Statistics and for generous mentorship of students and post-doctoral researchers".
She currently serves on the editorial boards of the Annals of Statistics, Journal of the American Statistical Association and Journal of the Royal Statistical Society (Series B). At the University of Chicago, she participates in several initiatives that promote the role of Statistics in society (Centre of Applied AI at Booth, Committee on Quantitative Methods in Social, Behavioral and Health Sciences, Data Science Institute). She is also an affiliate faculty at the Department of Statistics at the University of Chicago.
Beyond her academic pursuits, Veronika is an avid pianist, tennis enthusiast, and golf neophyte.