The Statistics Seminar speaker for Wednesday, October 25, 2017, will be Ben Shaby, an Assistant Professor of Statistics at Penn State University who develops statistical theory and methods to study extreme weather events. He works with climate scientists, hydrologists, and wildfire scientists in academia and government to understand and mitigate the risks associated with rare, high-impact events. Dr. Shaby was the recipient of the American Statistical Association Section on Statistics and the Environment's Early Career Investigator Award in 2016. He completed his Ph.D. at Cornell University in 2009 and held postdoctoral appointments at Duke University and UC Berkeley.
Talk: Hierarchical Scale Mixtures for Flexible Spatial Modeling
Abstract: Scale mixtures of Gaussian processes have emerged as desirable candidates for modeling extremal phenomena in space. They are intuitive, simple to describe constructively, and flexible in the types of extremal dependence that they can represent. Inferences for these models using censored likelihoods has been limited to very small datasets due to the presence of a high-dimensional Gaussian integral that must be evaluated numerically. Rather than integrating over a latent Gaussian process, we condition on it, expressing the model hierarchically. This way, we allow Markov chain Monte Carlo to do the hard integration, and open the door to inference on much larger datasets than were previously possible.