Research in statistics at the University of Bath covers a wide range of aspects related to handling uncertainty and is currently funded by EPSRC, MRC, Innovate UK, NIHR and Cancer Research UK.
Members of the statistics group work at the interface of methodology and application. We focus on producing models and methods which can be directly applied beyond academic statistics, and we are working with several national and international companies and organisations.
The applications we consider are varied and include health, environmental science and finance. In addition to core statistical research, members of the group undertake collaborative work where this has a substantial innovative component, and we are always interested in establishing new industrial partnerships and interdisciplinary projects.
The following summaries highlight our key areas of expertise.
Computational Statistics
Computational methods have become essential to use statistics in many real-world applications of statistics, due to the large amount of data or the complexity of the models. At Bath we have expertise in a range of methods to fit complex stochastic models. Emiko Dupont, Julian Faraway, Oliver Feng and Christian Rohrbeck are working in the areas of nonparametric regression and shape analysis, which have, for instance, applications in insurance and health. Kari Heine and Vangelis Evangelou are interested in developing Monte Carlo methods in a wide range of settings, such as biology and health. The statistics group is also a key member of the Centre of Mathematics and Algorithms for Data which brings together researchers in computer science, numerical analysis and machine learning.
Graphs, Networks and Time Series
High-dimensional and complex data structures occur in a wide range of applications, such as social media networks and image analysis. The research undertaken by Matt Nunes, Sandipan Roy, Simon Shaw and Vangelis Evangelou has led and continues to lead to new methodology and software to analyse data with a network structure, including graphical models. Matt Nunes research interests also include the analysis of time series data, and he is the Bath Lead of the NeST programme grant funded by EPSRC and led by Imperial College London.
Medical Statistics
We are highly involved in the application of statistics in medicine and the health sciences. Chris Jennison, Haiyan Zheng and Thomas Burnett are developing advanced clinical trial designs in collaboration with experts from the pharmaceutical industry. Haiyan Zheng currently holds a 6-year fellowship funded by Cancer Research UK. Theresa Smith's and Julian Faraway's research advances the application of statistics in a range of health applications, for instance, via their involvement in the Centre of Excellence in Water-Based Early-Warning Systems for Health Protection. Theresa Smith is also currently the theme lead for Healthcare Ecosystems in the AI for Collective Intelligence (AI4CI) research hub led by Bristol University. Karim Anaya-Izquierdo works in the area of survival analysis and further explores how design principles may be applied in engineering.
Spatial and Environmental Statistics
In many applications valuable insights can be gained by incorporating into the statistical analysis the spatial context in which, for instance, environmental and health data are usually collected. At Bath we are actively involved in the development of innovative statistical methodology for analysing spatial data. Emiko Dupont works on spatial regression models and spatial confounding. Vangelis Evangelou’s and Theresa Smiths' research interests are in the area of geostatistics, including the process of collection spatial data and their analysis. Christian Rohrbeck develops methodology which combines ideas from spatial statistics, non-parametric regression and extreme value analysis to analyse environmental extremes and their impacts.