Who writes this blog?
My name is Dr Ewan Cameron. I’m a Senior Computational Statistician in the Department of Zoology at the University of Oxford, working in the SEEG for the Malaria Atlas Project (MAP) to model the geospatial distribution of the global malaria burden. I’m an Australian, with a French/British wife, living in Britain. Apart from the fair city of Canberra, where I grew up, I’ve lived in St Andrews (Scotland; 4 years), Zurich (Switzerland; 2 years), Brisbane (Queensland; 2 years), and now Oxford (2
months years). One day I hope to own a mini-pig and a whippet. I now own a Bedlington terrier!
What are your qualifications?
A BSc (Hons) in Physics and Math from ANU and a PhD in Astronomy & Astrophysics from Mount Stromlo Observatory. Followed by three years of post-doctoral research in astronomy (St Andrews; ETH Zurich), two years in statistics at the Queensland University of Technology, and two years in epidemiological statistics at Oxford.
What do you research?
In astronomy my focus was the process of galaxy formation over cosmological time, as revealed by the comparison of near-field and high-redshift imaging surveys (esp. the SDSS, MGC, HUDF, and WFC3/ERS datasets). I stay involved through collaborations with the COIN (Cosmostatistics Initiative) project, attending the occasional workshop or mini-conference, and the Astrostats coffee group at Oxford.
In statistics at QUT my research concerned the theoretical and practical development of Bayesian methodologies, with a particular regard to problems in both astronomy (e.g. pioneering the use of Approximate Bayesian Computation for astronomical problem solving; and advancing parametric & semi-parametric model selection techniques (e.g. my Statistical Science paper) for testing the fine structure dipole hypothesis) and healthcare (e.g. inferring trends in incidence and prevalence of Chlamydia Trach. amongst the Australian population using routine national surveillance data).
In my current position in Zoology at Oxford I’m working on geospatial statistics for mapping the burden of P. Falciparum malaria infection around the world: see my [joint first author] Nature paper, my Nature Communications paper, and my profiles in Science and API News #selfpromotion. The key computational challenges we face in this endeavour concern efficient posterior sampling from large-scale, latent Gaussian random field models—again there is an interesting connection here to astronomical/cosmological data science (by way of the random field structure).
Why do you write this blog?
Having made the transition from astronomy to statistics (and now epidemiology) I am in the unique and fortunate position of being able to pass critical judgement on the state of astrostatistics as an outsider. While I may often lend a cynical voice to the discussion of astronomical research I do so with the hope of incrementally improving a field that I still hold dear, and may indeed one day return to (in the distant future). Perhaps I can aspire like Larry David in Curb Your Enthusiasm to be the first person to “eat where he once shat”.