Supervisor: Dr James Harte
The Institute of Digital Healthcare (IDH), WMG, University of Warwick is offering a 3-year studentship covering full support for tuition fees and an annual minimum tax-free stipend of Â£13,590 approx. The project is available to UK/EU nationals only due to the nature of the funding and will commence as soon as possible or at the latest Oct. 2013.
This project will explore the application of systems identification approaches in medicine and healthcare, and is part of a large EPSRC funded project aiming to develop a novel framework for patient-centred model-based predictive and diagnostic tools.Â The overarching goal of the project is to develop a hybrid of predictive learning from a repository of patientsâ data and biologically relevant personalized modelling for each patient.Â Data-centred machine learning methods typically aim to identify relevant patterns on scales larger than individual data items (patients).Â On the other hand, well-formed mechanistic models representing crucial biological aspects behind the patientâs data provide an opportunity to capture the underlying patientâs state more naturally than the crude measurements. A successful symbiosis of those two diverse perspectives will enable the end users (clinicians) to make informed decisions based on transparent and interpretable models, capturing general patterns detected across model sets underpinning individual patients.Â Â This EPSRC funded project will bring together academics and researchers from the Universities of Warwick, Birmingham, Bristol, University College London and Kingâs College London, working on early intervention in both physical and mental health conditions and modelling processes that accurately categorise patient health states.Â Â
Identification is the process of constructing a mathematical model of a system with unknown dynamics from observations and prior knowledge.Â System identification (SI) is an iterative procedure where an initial model structure is chosen (based on prior knowledge) or a general functional expansion of observable variables is used; coefficients are then fitted from the data using optimal estimation methods (i.e. Bayes; Max-likelihood; ordinary least-squares); and then the model structure is tested for redundant or missing terms, and the process repeated.Â This PhD project will explore the use of these approaches to medical and biological systems, where the challenge lies in dealing with uncertain data records or only approximate time scales.Â
This PhD project will require extensive mathematical ability, thus applicants should hold a minimum upper-second honours degree (or equivalent) or a Masterâs degree in a relevant subject such as mathematics, engineering, physics or computer science.Â Applicants with a demonstrable and strong interest in modelling biological systems will be favoured.
Awards available:Â 1 award available
Funding Details: Full support for tuition fees and an annual minimum tax-free stipend of Â£13,590 approx.
Length of Award: 3 years (PhD)
Eligibility:Â Due to funding restrictions this is available to Home (UK & EU) students
Any enquiries relating to the application process should be directed to Jennifer Kirkwood ([email protected]).
Any enquiries relating to the project should be directed to Dr James Harte atÂ [email protected].
Deadline:Â 30 June 2013