The Department of Chemical and Biomolecular Engineering welcomes Dr. Daniel Linder, an assistant professor of biostatistics and data science at the Medical College of Georgia, Augusta University, as a part of the ChBE 2018 Fall Seminar series. Dr. Linder’s research interests are in developing statistical methods for dynamical systems and Bayesian methods for high-dimensional data analysis. He is also interested in sampling techniques for study designs that improve statistical power.
His seminar, titled “Statistical inference in a class of biochemical reaction networks,” will take place on Thursday, September 20 from 2:00-3:00pm in Earle 100.
Statistical inference in stochastic dynamical systems from observed noisy trajectory data is notoriously difficult, primarily due to intractable likelihoods. Such inference methods often require sampling from the dynamical process at parameter values in an attempt to approximate the likelihood. In his seminar, Dr. Linder will discuss methods we have developed to tackle these kinds of challenges; for instance, ones based on algebraic statistics, regularization approaches and synthetic likelihood. Dr. Linder will present results from the method on real data sets, one from a regeneration experiment in the zebrafish, and if time permits, another from the plague outbreak at Eyam, Derbyshire England from 1665-1666.