Research Priorities

• Statistical inference - (1) how do we estimate location, scale and shape in distributions with heavy tails, (2) can we improve approximate Bayesian computation algorithms for models involving quantile distributions, and (3) how do we calibrate chain-binomial models using count data?
• Marine ecology - how do we estimate animal movements using information contained in family trees?

Researchers

• Chief Investigator: Phil Pollett
• Research Fellow: Ross McVinish
• PhD student: Robert Cope
• Honours student: Trent Spears

Collaborating Researchers

• Dr Olena Kravchuk, School of Agriculture, Food and Wine, University of Adelaide
• Dr Janet Lanyon, School of Biological Sciences, University of Queensland
• Dr Daniel Pagendam, CSIRO Division of Mathematics, Informatics and Statistics
• Dr Jennifer Seddon, School of Veterinary Science, University of Queensland
• Dr Christian Weiss, Department of Mathematics, Darmstadt University of Technology

Selected Research Projects

• Estimating Location, Scale and Shape in Long-tailed Distributions

  Project leader: Phil Pollett (UQ)
  Researcher: Olena Kravchuk (University of Adelaide)

  We have drawn on the relationship between the Cauchy distribution and the hyperbolic secant distribution to prove that the Maximum likelihood estimator of the scale parameter of the Cauchy distribution is log-normally distributed and to study the properties of a Hodges-Lehmann type estimator for the scale parameter. We have shown that the scale estimator is slightly biased, but performs well even on small samples independently of the location parameter.

  Research outputs

  • Kravchuk, O.Y. and P.K. Pollett (2012) Hodges-Lehmann scale estimator for Cauchy distribution. Communications in Statistics - Theory and Methods 41, 3621-3632.
• Approximate Bayesian Computation for Quantile Distributions

  Project leader and researcher: Ross McVinish (UQ)

  A new approximate Bayesian computation (ABC) algorithm is proposed specifically designed for models involving quantile distributions. The proposed algorithm compares favourably with two other ABC algorithms when applied to examples involving quantile distributions.

  Research outputs

  • McVinish, R. (2012) Improving ABC for quantile distributions. Statistics and Computing 22, 1199-1207.
• Optimal Design of Experimental Epidemics

  Project leader: Phil Pollett (UQ)
  Researcher: Daniel Pagendam (CSIRO Division of Mathematics, Informatics and Statistics)

  We consider the optimal design of controlled experimental epidemics or transmission experiments, whose purpose is to inform the practitioner about disease transmission and recovery rates. We demonstrate the broad applicability of our methodology using a diverse array of compartmental epidemic models.

  Research outputs

  • ¹ Pagendam, D.E. and P.K. Pollett (2013) Optimal design of experimental epidemics. Journal of Statistical Planning and Inference 143, 563-572.

  ¹ To be reported in connection with ARC Discovery Grant DP120102398 "Random Network Models with Applications in Biology".
  

• Parameter estimation in chain-binomial models

  Project leader: Phil Pollett (UQ)
  Researcher: Christian Weiss (Darmstadt University of Technology)

  We establish a connection between a class of chain-binomial models of use in ecology and epidemiology and binomial autoregressive (AR) processes. New results are obtained for the latter, including expressions for the lag-h conditional distribution and related quantities. We focus on two types of chain-binomial model of use in epidemiology and ecology, and present two approaches to parameter estimation. The asymptotic distributions of the resulting estimators are studied, as well as their finite-sample performance, and we give an application to real data. A connection is made with standard AR models, which also has implications for parameter estimation.

  Research outputs

  • ² Weiss, C.H. and P.K. Pollett (2012) Chain binomial models and binomial autoregressive processes. Biometrics 68, 815-824.

  ² To be reported in connection with ARC Discovery Grant DP110101929 "New Methods for Improving Active Adaptive Management in Biological Systems".

• Animal Movement Between Populations Deduced from Family Trees

  Project leader: Phil Pollett (UQ)
  Researchers: Robert Cope (UQ), Janet Lanyon (UQ) and Jennifer Seddon (UQ)

  We have develop new methods for estimating animal movements using information contained in family trees. Movement estimates are essential to population models that assist natural resource managers to plan species recovery and to predict the effect of future challenges, such as human-mediated activities and climate change. We have evaluated several methods for constructing family trees from genetic data and developed a statistic that describes animal movement between populations which is based on the families whose members were sampled in more than one population; empirical data has been sourced from a long-term mark-recapture study of dugongs in Moreton Bay, and new samples from two adjacent populations.

  Research outputs

  • Cope, R.C., Lanyon, J.M., Seddon, J.R. and P.K. Pollett (2012) Development and testing of a genetic marker based pedigree reconstruction system incorporating size-class data. Submitted for publication.