Can't display images!

PROFILE

John Donnelly is a senior consultant statistician in the CSIRO Division of Mathematics, Informatics and Statistics (CMIS), with training in agricultural science, biometrics and statistics from the University of Sydney and the University of New England, Armidale. He is an accredited statistician with the Statistical Society of Australia Inc. and is a member of the Australasian Region of the International Biometric Society.

John has been a consultant in biometrics and applied statistics for the past forty years, consulting with researchers in CSIRO Divisions and Australian Universities, government departments and companies in the business and finance sectors.

In 1970 John was appointed to the CSIRO Division of Mathematical Statistics, now CSIRO Mathematics, Informatics and Statistics (CMIS), and was employed at the CSIRO Pastoral Research Laboratory at Armidale to work on experimental designs and statistical analyses for large scale field experiments. After transferring to the Sydney office of DMS in 1975 John worked for about fourteen years as consultant statistician to the CSIRO Division of Animal Production at Prospect in areas such as repeated measurements analysis, experimental design, regression modelling and statistical software development for radio-immuno-assay.

From 1988 to 1999 John worked with the Signal Analysis Group in DMS and then the Analysis of Large and Complex Datasets Group in CMIS. He completed confidential projects for the Australian Petroleum Industries Research Association, the Sydney Water Board (now Sydney Water), Australian Water Technologies, David Jones (Australia) Pty Limited, the Environment Protection Authority of NSW, the Roads and Traffic Authority of NSW, Polartechnics, Gordon and Gotch Limited, Ross Carmichael Singer, the Institute of Respiratory Medicine University of Sydney, Britz: Australia, NSW TAB/Skychannel, Australian Taxation Office and Commonwealth Bank of Australia.

Other projects include:

  • predicting roof failure in coal mines using classification and regression trees (CART)
  • modelling laser Raman spectra profiles for the prediction of thermal maturity of coal leading to the identification of oil deposits
  • comparison of estimated spectral densities using nonparametric regression
  • determination of binary tree-based decision rules from ultrasound data for the indication of premature delivery to reduce neonatal mortality
  • the analysis of water quality data for the Sydney Waterboard Clean Waterways Program
  • the impact of the Sydney ocean outfalls on ocean water quality for the NSW Environment Protection Authority
  • the development of a computer software package for the analysis and display of a corporate confidence index
  • development of customised modules and on-line help for GSM, a statistical package providing sophisticated statistical methods in dynamic graphics and data analysis for the mineral industry
  • determination of confidence limits for the analysis of car exhaust fumes for the NSW Environment Protection Authority
  • a review of the accreditation and auditing procedure used by the RTA Southern Region for evaluating road condition
  • comparing variability in road roughness measurements collected by the RTA from a profilometer vehicle driven by drivers of varying skill level
  • using regularised discriminant analysis (RDA) to determine classification rules for an instrument being designed to identify patients with cervical cancer
  • comparison of special-purpose vehicles measuring skid resistance of roads at various sites around Sydney
  • methods of improving magazine supplies at newsagents and Supermarkets
  • comparison of new profilometer data from ARRB and the NSW RTA on a test circuit
  • statistical analyses for studies to evaluate short tests for motor vehicle emissions
  • statistical methods for the classification of airborne allergenic pollen and fungal particles
  • using statistical data mining to improve marketing strategies in the tourism industry
  • exploratory data analysis and sample size estimation for a study on exhaust emissions from diesel vehicles
  • modelling probability of whiplash injury in vehicles of different make and model
  • determination of pricing structures
  • identification of typical drive cycles for diesel vehicles
  • profile analysis to identify customers likely to adopt a new product
  • From 2000 to 2002 John was a member of the Business and Financial Engineering Group in CMIS, a CSIRO research and development group set up to explore opportunities for statistical data mining in business and finance. Confidential consulting projects were completed for major banks and insurance companies. Also, a statistical software package SSOFTRAC was developed for the NSW Roads and Traffic Authority to assist with automatic detection and correction of corrupt traffic counts.

    Following this John worked with the Health Data Integration Analytics Team for the Preventative Health Flagship in CSIRO to help build a prototype web-based workbench to integrate datasets from different sources and deliver Privacy Preserving Analytics (PPA) for the analysis of confidential data.

    John then helped set up the Qantitative Risk Management Group, which was required in CMIS to focus on developing methods for risk analysis and customised solutions for applications in industry. A new area of particular interest was the Basel II Accord and its implications for the banking industry. This led to research and development of methods for modelling frequency and severity of losses due to operational risk and collaboration with major Australian banks to build customised software systems for operational risk modelling to estimate regulatory capital required by Basel II.

    Following the introduction of matrix management into the organisational structure, the QRM group was merged into a new group called the Eneterprise Decision Making (EDM) group.

    Family

    Can't display images!

    Hobbies

  • Sport - cricket, footy, golf, tennis, snooker, etc.
  • Fishing
  • Music - just about anything except contemporary classical.
  • Photography - amateur only (Pentax ME Super user).

  • Frontierland

  • Maintained by John Donnelly
    Last updated on 20th August 2010


    Back to JD's home page


    © Copyright 2012, CSIRO Australia
    Use of this web site and information available from
    it is subject to our
    Legal Notice and Disclaimer and Privacy Statement