Gaussian Process Classification

Learning, detailing and experimenting with using Gaussian processes for classification. Exploring how to use approximations to avoid unnecessary computational slowdowns with an MCMC sampler.

Chicago Crime Classification

Modelling crimes in Chicago, using a public dataset. Logistic regression was used to predict whether an arrest would be successful or not, based on a number of predictors.

Downscaling Extremes of Precipitation in the South West

My Masters project, using extreme value theory to model maxima of precipitation across my home region. Comparing results from gridded model output and observations.