Date:  Thu, 26th March 2020

Time:
6:00pm - 6:30pm: AGM
6:30pm - 7.00pm: Refreshments
7:00pm - 8:00pm: H. O. Lancaster Lecture

Venue: New Law Building L024, The University of Sydney

Count on someone who can count! 

Count data occurs naturally in many different fields but they typically exhibit deviations, such as overdispersion, underdispersion, and zero-inflation, from the basic Poisson model. This means the Poisson distribution is often inadequate for modelling them. Incorrectly modelling such deviations leads to inefficient estimation and biased inferences on model parameters, resulting in invalid conclusions. In this talk, I will discuss some recent work in utilising the Conway-Maxwell-Poisson to handle over- and underdispersion and a new distribution generating mechanism to handle zero-inflation to obtain simpler and more easily-interpretable models. 

Biography
Thomas Fung is a Senior Lecturer in Statistics of the Department of Mathematics and Statistics at Macquarie University in Sydney. After serving on the NSW Branch of the Statistical Society as the Assistant Secretary and Secretary, he is now the incoming president. During his term, he hopes to further diversify the branch’s activities and create an even more inclusive environment. His research is motivated by data that exhibit deviations from an assumed model, such as dispersion, zero-inflation, skewness and heavy-tailedness.