Calibrating infectious disease models to real-time data remains a key computational and methodological challenge in outbreak response, representing a gap between the possibilities presented by such models and their use in the real world to inform policy. Bridging this gap is crucial to providing accurate and timely predictions of outbreak trajectories, evaluating control measures, and testing of control strategies.
In this project, the student will develop stochastic dynamical models and inference methodology, utilizing high performance computing techniques to calibrate these high-dimensional models in real time. Alongside model development, the candidate will create a modular software library for infection control. This software will enable users to design their own control strategy from a set of pre-defined options and assess their effectiveness. Strategies may include testing protocols, isolation measures and vaccination programmes.
The models and software will be applied in hospital settings, working with NHS partners to develop data-driven approaches for infection control. Hospital outbreaks are a major challenge for infection control and require targeted control measures to be implemented rapidly. Current models in these settings are often constrained by computational complexity and scalability as they need to explicitly represent the complex hospital environment, including ward layouts and the network of staff-patient interactions.
Objectives
- Construct individual-level stochastic dynamical models of transmission for different respiratory pathogens.
- Develop methodology to calibrate stochastic transmission models in real time.
- Create modular software to simulate and evaluate infection control strategies.
For more information, please see the job announcement.
