2nd year PhD student talks (see also 4th Dec)
Michael Thomson: Statistical Modelling of Equations of State
Rosanna Cassidy: Modelling and Analysis for Epigenetic Data
Jonathan Davies: Statistical methods for source localisation in Magnetoencephalography (MEG)
Abstracts
Michael Thompson
Statistical Modelling of Equations of State
Capturing and storing CO2 produced by power plants is a potential means to reduce emissions that contribute to climate change. This calls for a need to characterise, via an 'equation of state', the relationship between pressure, temperature and volume of CO2, and predict when phase transitions occur. I will discuss both parametric and non-parametric approaches to developing equation of state models. I will discuss the advantages and disadvantages of the two approaches and present numerical results comparing their predictive performance.
Rosanna Cassidy
Modelling and Analysis for Epigenetic Data
We discuss the inclusion of whole genome sequence data in models of disease transmission, with a particular focus on the spread of MRSA in hospital wards. We attempt to highlight some of the common limiting assumptions made in previous Bayesian models for such epigenetic data, to illustrate the impact that these assumptions have, and to develop new models. We look at an example of a real dataset.
Jonathan Davies
Statistical methods for source localisation in Magnetoencephalography (MEG)
MEG is a neuroimaging procedure that is based on the magnetic fields that result from neural activity. In source localisation problems we are interested in trying to infer the location of active brain areas. However, this is often challenging. Especially given the ill posed nature of the inverse problem. This talk will briefly outline the background of MEG and some of the challenges involved, before looking at some of the existing approaches to the problem. Finally I will present a few of the methods that I have begun to implement that provide some sparsity to the problem.