Statistical and Computational Modelling

Focus of our research

Complex systems in nature and technology often consist of many simple components, which combine to show complex collective behaviour. These systems require a combination of statistical, mathematical and computational modelling techniques to understand their behaviour. We use a combination of process-driven and data-driven modelling to develop powerful strategies to understand and predict these complex behaviours.

In addition to data-driven techniques, as embodied by machine learning, we use reverse engineering, as well as mathematical and statistical modelling techniques to analyse real-world problems in the bio-sciences and engineering. Whilst data-driven techniques are described as ‘model-free’, they are in fact ways of optimising model parameters, and the quality of the output is only as good as the underlying model.

Current projects for this theme:

• Understanding and predicting the onset of neurological illnesses such as Alzheimer’s and Parkinson’s disease
• Investigation of the psychological impact of flooding in the UK
• Spatio-temporal modelling of HIV, Cholera and COVID-19 for propagation, prediction and policy decisions
• Prediction of pedestrian behaviour to inform connected-vehicle control
• Modelling of protein dynamics during translocation across cell membranes