I am a mathematician and a statistician with a strong interest in biological systems, and I develop computational methods that help to uncover biological and epidemiological knowledge.

I direct Oxford's Schmidt AI in Science programmes: our Postdoctoral Fellowship programme, and our Visiting Faculty programme for Fellows from India and Africa. Both of these schemes provide funding for Fellows who apply methods from AI to advance scientific knowledge, and these schemes will collectively fund 164 Fellow-years of research by 2030.

I am a passionate communicator of statistical ideas (for example, through my YouTube videos), and I develop open-source materials which allow other researchers to understand cutting-edge methods.

Bayesian statistics, mathematical biology, epidemiology, Bayesian nonparametrics, research software

I joined the Statistics Department in September 2023 to work with David Steinsaltz on the development of new methods to study patients with multiple long-term conditions (“multi-morbidity”). The topic is closely related to my recent fellowship work that explored new ways to characterise and classify diseases (CTSU, Oxford Population Health, 2019-22), that was funded by the Nuffield Department of Population Health (NDPH). The fellowship work aimed to develop novel but rigorous methods for big-data epidemiological studies.  

I originally trained in theoretical physics with a Ph.D. on the ageing and stability of emulsions and foams supervised by Prof. Mike Cates at the University of Edinburgh. I joined UKAEA at Culham Science Centre in September 2000, and spent 15 years as a theoretical physicist studying plasma stability in nuclear fusion experiments. Following an M.Sc. in Applied Statistics at the University of Oxford in 2016, I joined NDPH, and my research was predominately based within Oxford University's Big Data Institute until September 2023. In addition to my fellowship project, I have worked on cancer epidemiology (CEU, 2016-2019) and statistical genetics (GSK-funded, 2022-2023). 

My main research interests involve the development and use of novel mathematical methods to utilise the information in large epidemiological datasets. My work has aimed to combine the best existing epidemiological methods with modern clustering techniques and biologically-motivated mathematical models of disease. This has included extensive multi-stage modelling of disease using UK Biobank data (21st Armitage Workshop: Talk by Dr Anthony Webster), and pioneering the use of the Poisson-Binomial model to quantify how much disease-risk can be explained by old age and established risk factors. This latter work provides a first attempt to quantify how much prior diseases are increasing future disease-risk in the UK Biobank cohort, above what is expected based on age and well-known risk factors. My previous research has ranged from theoretical results in soft condensed matter and plasma physics, to cancer epidemiology and climate change. Broadly, my interests lie in the development and use of mathematical and statistical models for problems with social or public health value.

Contact me for:

• Support with StatML CDT related matters
• Support with student finances

My background is at the intersection of research and practice in Responsible Artificial Intelligence. I previously worked with AI startups, accelerators, research institute and government bodies on multistakeholder projects and initiatives pertaining to RAI.

I am a Florence Nightingale Bicentenary Research Fellow in the Department of Statistics at the University of Oxford. Prior to this, I completed my PhD at Imperial College London under the supervision of Professor Heather Battey.

My research lies in theoretical statistics, where I aim to develop methods for inference with guarantees in high-dimensional and non-parametric settings. For access to code that implements these methods, please see rm-lewis (github.com).