Research interests
The main research interests of the Department fall into following categories.
Graphical models and their applications
Bayesian statistics
Computationally-intensive methods
Statistical genetics
Bioinformatics
Applied probability
Discrete mathematics and operational research
Applied statistics
There is substantial overlap and interaction between researchers in the department, particularly in the development of computationally intensive statistical methods, such as MCMC and sequential Monte Carlo techniques for Bayesian inference.
Graphical models
Professor Steffen Lauritzen is an internationally recognised expert on graphical models, with a distinguished publication record in this area. Graphical models play an important role in understanding the relationship between variables in complex datasets. They contribute to statistical understanding of real world phenomena such us medical diagnosis and business planning while posing substantial methodological and computational challenges.
Bayesian and computationally-intensive methods
Dr Peter Clifford, Dr Geoff Nicholls, and Dr Chris Holmes work in this general area, focusing on problems associated with high-dimensional parameter spaces and extensive and evolving data sets. Professor Brian Ripley has a particular interest in finding and visualising structure in large complex data sets. This has applications for example to genetics, insurance databases, and brain imaging.
Applied probability
Measure-valued diffusions and genealogical processes arise naturally from the models of population evolution considered by the mathematical genetics group. At an abstract level, challenging problems arise in understanding the structural properties of the processes. Other areas of research in applied probability include describing spatial mosaics (patterns of colours in two-dimensional images) and their extensions to higher dimensions. Some members of the department are part of interdepartmental groups with interests in mathematical finance. Stochastic processes and differential equations are major topics in this context stimulating research projects in a more theoretical framework.
Discrete mathematics and operational research
Research interests in the Department include combinatorics, stochastic scheduling, and portfolio planning. There is a particular focus on applications to the management of pharmaceutical research and to radio channel assignment. There is also interest in the stochastic modelling and optimization of networks, with applications to computer and communication systems.
Mathematical genetics and bioinformatics
Professor Peter Donnelly and Professor Jotun Hein lead research groups working in mathematical genetics and bioinformatics.
They are concerned with the development and application of models
and statistical methodology for the analysis of data from modern
genetics. Major themes include coalescent models, computationally
intensive inference in molecular population genetics, human variation
and disease, molecular evolution, protein structure prediction,
stochastic grammars, statistical alignment, and sequence analysis.
Applications of this modelling to specific genetic systems use
analytical and computational methods to further our understanding of
evolution, both quantitatively, and through the development of
efficient statistical methods for analysing molecular data. One
particular area of interest in the department is in using data from
diverse modern populations to shed light on early human evolution. A
related interest is in the statistical issues arising from the use of
DNA profiles in criminal trials.
Further information
List of possible research areas for entry in October 2010
Departmental research pages
Full list of academic staff
Printer Friendly Version 