My core research is in computational statistics. I develop and analyse algorithms for Bayesian inference, including sequential Monte Carlo and Markov chain Monte Carlo. I work with applied scientists to develop principled and robust data analysis for their data. Often this involves designing Bayesian models to incorporate domain and expert knowledge. I also design open source software for statisticians, scientists and industry.

Since September 2023, I am a postdoctoral researcher supervised by Professor Christian Robert at Université Paris-Dauphine, Paris Sciences & Lettres. I am working on principled Bayesian inference algorithms with data privacy guarantees. In pursuit of this goal we are developing a framework for defining privacy guarantees for use in scientific, industrial and other contexts with differing privacy requirements.

In Paris I co-organise the Mostly Monte Carlo Seminar Series. In 2023 I was chair of the committe organising SMC Down Under.

Current projects

• A Bayesian decision-theoretic framework for statistical data privacy
• Monte Carlo twisted particle filters arXiv
• Calibrating surrogates models for intractable Bayesian posteriors arXiv
• Particle filters for binary state spaces
• Tidy software for BART causal treatment effect models R package

Selected Journal articles

• “Accelerating sequential Monte Carlo with surrogate likelihoods” (2021) Statistics and Computing https://doi.org/10.1007/s11222-021-10036-4 with A Lee, and C Drovandi.
• “Analysing Sensitive Data from Dynamically-Generated Overlapping Contingency Tables” (2020). Journal of Official Statistics https://doi.org/10.2478/jos-2020-0015 with B Baffour, M Spallek and M Haynes.
• “Polling bias and undecided voter allocations: US Presidential elections, 2004 - 2016” (2019). Journal of the Royal Statistical Society (Series A) https://dx.doi.org/10.1111/rssa.12414 with T Ballard and B Baffour.
• “Fitting monotone polynomials in mixed effects models” (2019). Statistics and Computing https://doi.org/10.1007/s11222-017-9797-8 with BA Turlach and K Murray.

See Google Scholar for more details.