Robert King's Research

Current research interests

Flexibly shaped distributions

The Generalised lambda distribution and the g-and-h and g-and-k distributions are families of distributions that allow a great variety of shapes within one distributional form. I am interested in fitting these distributions to univariate data and developing modelling methods using them. (fitting methods: starship; modelling: starship regression)

Bayesian Belief Updating in Robotics

Localisation and vision interpretation using efficient Bayesian computation.

Assessing Ecological Models

Theoretical ecology has produced a large number of models. I am interested in methods for assessing the closeness of such models to observed environmental data. (for sex-ratios and biomass apportionment)

Socio-Spatial Statistics

Bringing together multi-variate and spatial statistical methods to characterise social structure. Applied to such problems as characterising disadvantage in a particular area. This work includes collaborative research with NSW government departments in the Hunter Region and has also led to work on the processes required for data sharing with government agencies.

Computing for Statistics

Computational methods for statistics, particularly simulation-based methods. I maintain the Newcastle archive of the R mailing lists, R-help, R-devel, R-announce, R-packages and R-downunder.

Statistics for Computing

Applying statistical methods to investigation of the software engineering process.

Other Applied Statistics

Including in Readability and Anaesthetics.

Publications

• Airborne laser scanning: Exploratory data analysis indicates potential variables for classification of individual trees or forest stands according to species, Moffiet Trevor Noel, Mengersen K, Witte C, King Robert Arthur, Denham R, ISPRS Journal of Photogrammetry and Remote Sensing, 59 289-309 (2005)
• An Assessment of the randomization test for sex ratio biases, John G. Ewen, Phillip Cassey, and Robert A. R. King, The Auk, 120 (1) 62-68, 2003.
• The problem of testing the goodness-of-fit of stochastic resource apportionment models, Phillip Cassey and Robert A. R. King. Environmetrics, 12(3):691-698, 2001.
• A starship fitting method for the generalized lambda distribution, Robert A. R. King and H. L. MacGillivray, Australia and New Zealand Journal of Statistics (1999) 41(3), pages 353--374. The C programs for the starship method as used in this paper are available here, however, the R implementation is easier to use, and contains a number of numerical improvements on the version used in the paper, so I would recommend it instead.

Conference Presentations

• Starship Regression: A Regression Method with Flexibly-Shaped Errors. Conference Handbook, IBC, in parallel with ASC 2004, July 2004 Cairns, Qld
• What can we predict for the establishment success of novelty land bird introductions?,Statistics in Public Resources and Utilities and in Care of the Environment (SPRUCE), June 2003, Lund
• A monte carlo method for assessing the quality of a logistic based binomial estimate: presented at the 16th Australian Statistical Conference, Canberra, July 2002
• Characterising the shape of the generalised lambda distribution: presented at the Australasian Biometrics and New Zealand Statistical Association Joint Conference, Christchurch, December 2001
• Goodness-of-fit methods for species abundance distributions (King and Cassey): Presented at the 15th Australian Statistical Conference, Adelaide, July 2000
• Comparing biomass distributions (King and Cassey): presented at the Biometrics `99 conference.
• Are things getting more extreme these days? - detecting changes in tail properties - presented at Biometrics '97 (Wirrina Cove, South Australia, December 1997)
• Approximating distributions using the generalised lambda distribution - presented at the Sydney International Statistical Congress (1996)

Software

Here is some free software to implement some of these methods.

Publications information	Robert King's page	Statistics	Mathematical and Physical Sciences	University of Newcastle