Researcher

Emeritus Professor William T M Dunsmuir

My Expertise

Statistical modelling & analysis; time series methods; spatio-temporal statistical models; applications to meteorology & other environmental measurements in space & time; applications to time series in public health

Fields of Research (FoR)

Statistical Theory, Applied Statistics, Econometric and Statistical Methods, Financial Econometrics

SEO tags

Biography

 

Professor of Statistics at the University of New South Wales (1993-2019)  and Head of the Department of Statistics (1993-2000; 2003-2005, 2007). Previous academic positions: the Massachusetts Institute of Technology (1976-1980), The Australian National University (1980-1982), The Australian Graduate School of Management (1987-1988), Bond University (1988-1993) and the University of Minnesota (2001-2002). Previous consulting management positions:...view more

 

Professor of Statistics at the University of New South Wales (1993-2019)  and Head of the Department of Statistics (1993-2000; 2003-2005, 2007). Previous academic positions: the Massachusetts Institute of Technology (1976-1980), The Australian National University (1980-1982), The Australian Graduate School of Management (1987-1988), Bond University (1988-1993) and the University of Minnesota (2001-2002). Previous consulting management positions: Siromath Pty Ltd (1982-1987) and Bond University.

Education

  • Ph.D. (Statistics) 1977, Australian National University
  • B.Sc.(Hons) 1973, La Trobe University
  • Fellowship Diploma (Mathematics) 1971, R.M.I.T. University.

Professional affiliations:

I am a member of several professional associations: American Statistical Association (Fellow), Bernoulli Society, Institute of Mathematical Statistics, International Statistical Institute (Elected 1993), Statistical Society of Australia.

Consulting to Industry, Commerce and Government:

Commercial consulting has proved invaluable in informing my teaching and research in statistical methods. I developed substantial experience as a statistical consultant through five years full time as a consulting manager and consultant for Siromath Pty Ltd, a mathematical, statistical and computing consultancy established by CSIRO in 1980. Additionally while at Bond University I was Director of the Centre for Quality Management and Decision Analysis for five years. Since being at UNSW I have provided consulting and contract research activities on major projects including more recently on modelling policy impacts on criminal charge and conviction rates, impacts of government education programs on prescription drug utilization, time series and forecasting and assessment of prudential regulation risk management. Additionally, I have provided advice to various leading law firms and their clients on the use of statistical analysis in legal matters.

Awards:

  • Honorary Life Member, Statistical Society of Australia (2016).
  • Service Award, Statistical Society of Australia (2012).
  • Fellow: American Statistical Association (2004). Citation: For fundamental contributions to the theory and practice of time series analysis and spatio-temporal modelling, for excellence in teaching, administration and consulting and for service to the profession.
  • The Tjalling C. Koopmans Econometric Theory Prize with Richard A. Davis, 1998, for the paper “Maximum Likelihood Estimation for MA(1) Processes with a Root on or Near the Unit Circle,” Econometric Theory, Vol 12, 1996, pages 1–29.
  • D.M. Myers University Medal, La Trobe University, 1973.

My Grants

Inference for Hawkes processes with challenging data. ARC Discovery Grant 2024-2026 --Chief Investigators (University of NSW) Feng Chen, Thomas Stindl, William Dunsmuir. Partner Investigator (The Institute of Statistical Mathematics, Japan) Jiancang Zhuang.


My Research Activities

My core research has been in statistical time series analysis with excursions into spatio-temporal processes, longitudinal data analysis and applied statistics in a range of disciplines.

Currently my research is concerned with:

  • Methods and models for discrete valued time series.
  • Self exciting point processes.

Research in Detail

My theory and methods research has been into asymptotic theory for time series model estimation, methods for missing data in time series analysis, robust and non-parametric estimation in time series and image analysis, modelling seasonally varying time series, Bayesian methods for hierarchical spatial modelling and, more recently, modelling nonlinear non-Gaussian time series particularly focussed on discrete valued time series and, most recently, inference for Hawkes self-exciting point processes.

I have also engaged in substantial applied research with research collaborators in a variety of disciplines including public health, medical research and weather and environmental modelling.

View less

Location

School of Mathematics & Statistics
University of New South Wales
Sydney, NSW, 2052
The Red Centre
Room 2057





Map reference (Google map)