About the Project
A fully funded PhD opportunity to participate in the world-leading research undertaken by the EPSRC Doctoral Landscape Award at the University of Sheffield.
Supervised by Prof George Panoutsos and Dr Inaki Esnaola.
The University of Sheffield is one of the leading Russell Group universities in the UK. We carry out cutting-edge research with strong links to industry. When you enrol to do a PhD with us, you will be working with world-leading academics and have access to top of the range facilities. As a PhD student you will have the opportunity to gain skills not only to conduct research, but also to take your career to the next level, whether you want to stay in academia, go into industry or the public sector, or set up your own company. You will have access to a range of training and support services to help you excel in your studies and beyond.
Project Description
In recent years we have observed tremendous growth in the engineering applications of machine learning to accelerate and assist tasks in a range of industry sectors. This growth has also revealed a fundamental challenge and a barrier in the use of such machine learning tools in critical sectors. The challenge of 'trustworthiness'. The narrative in machine learning is gradually shifting from 'striving to be faster/more accurate/more sophisticated' to 'do we understand sufficiently, hence trust, decisions and outcomes derived from machine learning?'. A central challenge is therefore to move beyond empirical validation and provide quantitative measures of predictive reliability. In particular, theoretical generalisation bounds can describe how well a learned model is expected to perform on unseen data, but their practical value depends on whether the associated quantities can be estimated accurately and efficiently. In the context of engineering applications, this could relate to, for example, simple predictive tasks, such as regression and classification. For critical applications (e.g. related to human safety, defence and security), not being able to guarantee outcomes is a barrier to adoption. Aerospace parts manufactured with the assistance of machine learning cannot be certified for flight, nor can healthcare decision-making algorithms be qualified for use in clinical practice, unless full accountability and robustness of outcomes and machine learning performance can be guaranteed.
Statistical learning theory, leveraging information-theoretic tools, has provided, to an extent, robust theoretical bounds and generalisation guarantees for such machine learning models. The computational evaluation of such bounds, as well as sampling from the associated complex distributions, remains a severe bottleneck (as these deem the associated methods as impractical to use). It is timely to conduct research on these topics as success would unlock the use of a class of machine learning tools in critical applications, yielding both significance and impact.
The aim of this project, in collaboration with Wayland Additive, is to develop efficient Langevin-dynamics-based sampling methods for estimating the quantities required by information-theoretic generalisation bounds (Years 1-2). The successful development of this framework will be subsequently applied to empirically validate these theoretical bounds in real-world, industry-relevant case studies (Years 2-3) in advanced manufacturing. We expect the research to yield world-leading outputs, contributing to machine learning and its use in engineering applications more broadly, while paving the way for impact in collaboration with our industrial partners.
Candidate profile
We welcome applicants with a strong background in mathematics, statistics, computer science, electrical and electronic engineering, or a related discipline. The project will particularly suit a candidate with interests in probability, optimisation, machine learning theory, Bayesian inference, or uncertainty quantification. Strong programming skills, ideally in Python, are expected; prior experience with stochastic sampling, information theory, or machine learning research is desirable but not essential
How to apply
Interested candidates are strongly encouraged to contact the project supervisors to discuss your interest in and suitability for the project prior to submitting your application.
Please refer to the EPSRC DLA webpage for detailed information about the EPSRC DLA and how to apply.
Applicants must ensure that they use the correct application form linked to the EPSRC DLA webpage and must submit an application for October 2027 entry. The form is unique to the EPSRC DLA. It contains the list of advertised projects from which to select, and three questions specific to the DLA.
The application deadline is 17:00 GMT 13 January 2027.
The majority of our DLA studentships will be allocated to candidates who qualify for Home fee status, in accordance with EPSRC funding regulations which place strict limits on international recruitment.
Funding Notes
The award will fund the full tuition fee and a maintenance stipend at the UKRI rate (currently £21,805 per annum) for 3.5 years, as well as a research grant to support costs associated with the project.