Research

I am broadly interested in computational statistics and machine learning methodologies, and their applications in scientific and health-related disciplines In particular, I am interested in (1) developing methodologies with strong relevance to applications, and (2) advancing our understandings of the underlying mechanisms behind existing, commonly used algorithms. Below are some of the keywords that interest me:

• MCMC, SMC, Variational Bayes
• Gaussian Processes
• Bayesian Optimisation, Active Learning
• Diffusion Models, Computational Optimal Transport

Currently, I am investigating how Bayesian optimisation and active learning, in combination with physics-informed Gaussian processes surrogate models, can help engineers and oceanographers to better understand ocean currents and inform their exploration strategies.

Publication

Preprint

• Livingstone, S., Nüsken, N., Vasdekis, G., & Zhang, R. Y. (2024). Skew-symmetric schemes for stochastic differential equations with non-Lipschitz drift: an unadjusted Barker algorithm. [arxiv] (submitted)