Mathematics Colloquium: Professor Peter Green
Topic: Graphs, probabilities, data, models and learning
This talk will begin with an easy introduction to the role of graph theory in describing the qualitative relationships between variables, in both probability and statistics. No prior knowledge of anything is assumed! It turns out that graphs have a role to play in almost all aspects of modern model-building and data analysis, in modelling, elicitation, exposition, presentation and even computational algorithms. The area is a satisfying mix of graph theory, algorithms, probability and statistics. Later in the talk, I will say a little about some of my own contributions to the field, including some results in graph theory and on distributions of random graphs, and novel stochastic algorithms for model-fitting.
About the speaker
Professor Peter Green (UTS and University of Bristol)
Professor Peter Green, Fellow of the Royal Society, is a Distinguished Professor of Statistics, and also holds a Professorial Research Fellowship at the University of Bristol in the United Kingdom. His current research focus is on graphical models, forensic statistics, and theoretical properties of Bayesian procedures for inverse problems. He has a long-standing interest in many aspects of Bayesian analysis, especially modelling and computational issues. His research is described in full on his webpage.