Taylor Fry

 

Hugh Miller

Senior Actuary

About Hugh

Since joining Taylor Fry in 2005 Hugh has built a reputation as a reliable and thoughtful actuary working across a wide range of actuarial and analytics projects.

Hugh loves projects at the cutting edge of actuarial science where he can make use of his expertise in statistics to propose and implement new solutions to problems.

Hugh’s particular interest is “big data” projects in insurance and other analytics areas. He recently developed sophisticated general insurance pricing models and customer profiling models in telecommunications.

Hugh is an active and respected researcher whose awards include the 2011 University of Melbourne’s Chancellor’s Award for his PhD thesis on Statistical Methods for the analysis of high-dimensional data.  His 2010 paper Towards a better inflation forecast won the 2011 A M Parker Prize for the most outstanding paper published by the Actuaries Institute. He has also been awarded a number of actuarial education prizes, including the 2013 Andrew Prescott Memorial award for best overall performance in the Australian Fellowship Examinations.

Experience

Since joining Taylor Fry, Hugh has undertaken both traditional actuarial work and innovative analytics work in a variety of areas.

Social welfare valuation and analytics

  • Hugh has been a core member of the Taylor Fry team that performed annual valuations of the New Zealand Welfare system, the first valuation of its kind anywhere in the world. The project required significant statistical and computational expertise to provide analytical insight to drive future policy and programs.

Insurance Liability Valuation

  • Hugh has undertaken valuations for a wide variety of companies and schemes, including Workers’ compensation schemes, CTP schemes, Multi-line commercial insurers, Reinsurers and Specialist Consumer credit insurers.
  • Hugh has also used his statistical modelling skills to build statistical case estimate models for a workers’ compensation scheme, providing automated monitoring and improved accuracy in case estimation.

Customer Insights

Hugh has also been involved with a number of analytics projects involving data preparation, predictive modelling and model interpretation:

  • Assessing the accuracy of Australian Commonwealth Government welfare benefits and determining the best allocation of resources to review customers
  • Understanding customer product choice and price elasticity for a large telecommunications company

GI Technical Pricing and Optimisation

  • In particular, Hugh has applied his detailed knowledge of analytical and “big data” techniques to the risk pricing of personal lines in general insurance, resulting in optimal accuracy of pricing for his client.

Publications

  • Using generalised correlation to effect variable selection in very high dimensional problems by Peter Hall and Hugh Miller. The Journal of Computational and Graphical Statistics, 18 533-550, 2009.
  • Using the bootstrap to quantify the authority of an empirical ranking by Peter Hall and Hugh Miller. The Annals of Statistics, 37 3929-3959, 2009.
  • Predicting customer behaviour: The University of Melbourne’s KDD Cup report. By Hugh Miller, S Clarke, S Lane, A Lonie, D Lazaridis, S Petrovski and O Jones. JMLR: Workshop and conference Proceedings, 7 45-55, 2009.
  • Towards a better inflation forecast by Hugh Miller. Presented to General Insurance Seminar of the Institute of Actuaries of Australia 2010. (Winner of the 2011 A M Parker Prize for the most outstanding paper published by the Actuaries Institute.)
  • The rise and rise of hybrid modelling by Hugh Miller. Presented to General Insurance Seminar of the Institute of Actuaries of Australia 2010.
  • Statistical methods for the analysis of high-dimensional data by Hugh Miller. PhD Thesis, University of Melbourne Library 2010.
  • Local polynomial regression and variable selection by Hugh Miller and Peter Hall. Borrowing Strength: Theory Powering Applications – A Festschrift for Lawrence D. Brown, 216-233, 2010.
  • Sequential, bottom-up variable selection for high-dimensional classification by Peter Hall and Hugh Miller. Australian and New Zealand Journal of Statistics, 52 403-421, 2010.
  • Bootstrap confidence intervals and hypothesis tests for extrema of parameters by Peter Hall and Hugh Miller. Biometrika 97 881-892, 2010.
  • Modelling the variability of rankings by Peter Hall and Hugh Miller. The Annals of Statistics, 38 2652-2677, 2010.
  • Determining and depicting relationships among components in high-dimensional variable selection by Peter Hall and Hugh Miller. The Journal of Computational and Graphical Statistics, to appear, 2011.
  • A topic of interest: how to extrapolate the yield curve, by Peter Mulquiney and Hugh Miller. Presented at the Actuaries Institute General Insurance Seminar, 2012.
  • Discount Rates in General Insurance Pricing, by Peter Mulquiney, Brett Riley, Hugh Miller and Tim Jeffrey. Presented at the Actuaries Institute General Insurance Seminar, 2014.

CONTACT Hugh

Hugh.Miller@taylorfry.com.au

(02) 9249 2926 | Sydney

 

Areas of expertise

  • Insurance Liability Valuation
  • Customer Insights
  • GI Technical Pricing and Optimisation

Qualifications

  • Fellow of the Institute of Actuaries of Australia since 2013
  • Doctor of Philosophy in Statistics, The University of Melbourne (2011)
  • Bachelor of Science (Honours), The University of Sydney (2005)