Fast Solvers for Applications Involving Gaussian Processes

  • Studentship
  • Glasgow
  • 15 August 2022

Website UniStrathclyde The University of Strathclyde

  • Postgraduate research opportunities
  • Fast Solvers for Applications Involving Gaussian Processes
  • Mathematics and statistics
  • Duration: 3 years
  • Funding: The studentship covers full tuition fees and a tax-free stipend for 3 years
  • Starting date: starting in October 2022

About the Project

Gaussian Process (GP) regression is a powerful means of modelling correlated effects in many applications, including material forming, air pollution, instrument drift in metrology, and energy demand. Gaussian processes also underpin certain machine learning algorithms. GPs learn this correlation structure from measurement data and automatically allow for uncertainty quantification, making them attractive in industrial settings. Typically, fitting GP regression models requires the solution of one or more linear systems, which requires efficient linear algebra tools for both large- and small-scale problems. This project will develop the theoretical framework for solving these GP linear systems and will apply these to applications in advanced manufacturing in conjunction with the Advanced Forming Research Centre (AFRC).

This is a very exciting project which will allow the successful candidate to work at the interface of computational mathematics and machine learning, with a strong potential for industrial application. This inter-disciplinary research will be undertaken primarily within the Numerical Analysis and Scientific Computing group in the Department of Mathematics and Statistics at the University of Strathclyde. There will be frequent interactions with the AFRC and international partners. The studentship is part of a joint PhD Cluster with TU Delft, which will present unique opportunities for collaboration.

What’s in it for you/Benefits:

The successful candidate will attend regular research seminars and events within the Department of Mathematics and Statistics and will interact with a multi-disciplinary team and international collaborators. They will have opportunities to participate in conferences and workshops, and develop both technically and professionally.

Eligibility:

The ideal candidate will have a strong background in some of the following areas: linear algebra, numerical analysis, optimisation and/or machine learning. Experience in programming (e.g. MATLAB, Python or Julia) is highly desirable.

Applicants should have, or be expecting to obtain in the near future, a first class or good 2.1 honours degree (or equivalent) in mathematics or a mathematical science. An MMath, MPhys or MSc degree is desirable. To be eligible for a full award a student must have no restrictions on how long they can stay in the UK and have been ordinarily resident in the UK for at least 3 years prior to the start of the studentship. Deadline for application: 15 August, 2022.

How to apply:

Interested candidates are strongly encouraged to contact the first supervisor for informal discussions about the project.

Funding Notes

Fully funded 36-month PhD studentship in the Department of Mathematics and Statistics at the University of Strathclyde, Glasgow, United Kingdom, jointly supervised by Dr Jennifer Pestana and Prof. Victorita Dolean (Mathematics and Statistics) and Dr Salah Rahimi (AFRC).

The studentship covers full tuition fees and a tax-free stipend for 3 years starting on 1 October.

The University of Strathclyde is committed to achieving and promoting equality of opportunity in its learning, teaching, research and working environments, and to ensuring these environments support positive relations between people, and a culture of respect. As a provider of employment and education, we value the diversity of our staff and students and are committed to encouraging everyone to realise their full potential.