Kaiyu recently completed her PhD in Statistical Science in the department under the supervision of Prof. Serge Guillas and Dr. François-Xavier Briol, with a thesis called “Multilevel Methods for Monte Carlo Integration, with Applications to Tsunami Modelling”. Tsunami modelling is typically consists of using a large number of expensive simulations from systems of differential equations describing the evolution of her wave to estimate quantities of interest for risk estimation. Her thesis develops novel multilevel Monte Carlo methods based on Gaussian process models which can estimate these important quantities of interest much more efficiently than existing approaches.
Link to the MAPS news: https://www.ucl.ac.uk/mathematical-physical-sciences/news/2024/aug/maps-2024-postgraduate-phd-innovation-enterprise-student-prize-competition-winner