Before joining the Compass CDT, I was awarded First-Class honours in Mathematics from The University of Edinburgh and completed an MSc in Data Analytics at Queen’s University Belfast, with Distinction. I have also completed an internship with Danske Bank where I applied machine learning techniques to aid with the protection of vulnerable customers.
My PhD is in the area of Uncertainty Quantification and involves the emulation of complex computer codes which constitute a ‘black-box’ approximation of a physical experiment. Ultimately, we are interested in using the computer code within some form of decision-making process. Hence, for decisions made using the output of the code to be credible, it is necessary to approximate the uncertainty associated with the code. We are primarily interested in deterministic, expensive computer codes, for example finite element simulators run over fine meshes, or climate models. Emulation allows us to use a statistical model to approximate our computer code and provide an estimate of the associated uncertainty. Our current focus is on a particular type of emulator called a Gaussian process. Specifically, we are interested in using Gaussian processes to estimate functions for which there exists prior knowledge of shape constraints, such as monotonicity. This project is supervised by Professor Oliver Johnson.
Take a look at Conor’s blog post explaining uncertainty quantification methods for expensive computer experiments.