DPhil, Big Data Institute
Project: Modern machine learning methods
Cornelius obtained his BSc in Psychology from the University of Groningen, where he focused on research methods from early on. The crisis in replicability in the social sciences has raised his interest in Bayesian methods which soon became his primary focus. He went on to do a minor in theoretical statistics at the Chinese University of Hong Kong and obtained his MSc in Statistics with distinction from Warwick University in 2018. During his masters, he focused on Bayesian methods, classical machine learning methods, as well as, stochastic simulation methods, in particular MCMC. For his master dissertation he looked at high-dimensional heterogeneous socioeconomic and biological data to predict perinatal depression using various machine learning methods. In particular, Bayesian variable selection effectively identified relevant features from a sparse feature space which further fuelled his interest in modern Bayesian methods. Having a background both in social as well as mathematical sciences he recognises the value of interdisciplinary learning.
About Cornelius’ research
Cornelius’ research interests are directed towards modern machine learning methods, which are inspired by real world mechanism that we want to model.
Having a background in statistics he prefers a more theoretical approach to machine learning focusing on model building and formal reasoning. He is funded by Cancer Research UK. Having been in touch with several researchers from the Oxford Cancer Centre, he has gained an insight into some of the exciting research being conducted in Oxford.