Director of Studies: Dr David Walker
2nd supervisor Dr Alma Rahat
3rd supervisor Dr Matthew Craven
Applications are invited for a three-year PhD studentship. The studentship will start on 1 October 2019.
Optimisation problems abound in science and industry, and in recent decades a plethora of approaches have arisen to solve them – many of these approaches fall under the umbrella term of evolutionary computation (EC), a branch of artificial intelligence. As well as generating good solutions to optimisation problems, it is important that the processes with which they are generated are understandable by non-expert problem owners, and often this is not the case. Visualisation is a natural approach to addressing this issue, exposing the solutions and mechanisms used to generate them to the end user.
This project will seek to identify methods for visualising solution generation mechanisms used within optimisation algorithms. Well known examples are crossover and mutation used within genetic algorithms, however this work will explore a wider range of evolutionary algorithms (EAs), hyper-heuristics and Bayesian optimisers, visualising the solution mechanisms used therein. Visualisation in these areas to date has mostly considered the visualisation of solution sets, so this work is an exciting opportunity to advance knowledge and practice in this area. As well as developing visualisations, the project will involve validating them through usability studies with real end users.
The student will work in the School of Computing, Electronics and Mathematics in a highly interdisciplinary environment, working on the interface between computer science and mathematics. Beyond academia the project will involve engaging with industrial optimisation problems, which will enable the student to develop a wide range of skills.