Robert L. Wolpert ('72) is the 2019 Cornell Distinguished Alum Award recipient. He is a professor of statistical science at Duke University.
From Wolpert's faculty page: I'm a stochastic modeler-- I build computer-resident mathematical models for complex systems, and invent and program numerical algorithms for making inference from the models. Usually this involves predicting things that haven't been measured (yet). Always it involves managing uncertainty and making good decisions when some of the information we'd need to be fully comfortable in our decision-making is unknown.
Talk: Lévy-based Nonparametric Bayesian Models and their Applications
Abstract: We teach our students about ARMA and diffusion models with nicely-behaved sample paths and tame tail behavior, and then send them out into a world where they encounter time series and stochastic processes that feature missing data, jumps, heavy tails, and non-stationarity.
This talk presents and applies an approach to modeling time series, random processes, and random fields based on stochastic integrals with respect to Lévy random fields that may feature jumps, fat tailed distributions, and irregular sampling. Closed-form expressions are difficult to find, but simulation-based Bayesian posterior distributions are available. Motivating applications are drawn from the fields of proteomics, epidemiology, volcanology, and especially astrophysics.