Multi-objective optimisation refers to an optimisation scenario having more than one objective to be considered simultaneously. It exists ubiquitously. When we travel, we care about the time and the cost. When buying a car, we care about its performance and price. A major difference of multi-objective optimisation to single-objective optimisation (i.e., global optimisation) is that there exists no single optimal solution that can achieve the best for all the objectives, but rather a set of tradeoff solutions (called Pareto optimal front) which are not comparable to each other. Evolutionary algorithms have been found to be well suited to multi-objective optimisation. Its population-based search can provide a good approximation of the Pareto optimal front, with each individual representing a unique tradeoff between objectives.
Centred around evolutionary multi-objective optimisation (EMO), two pivotal issues are solution generation and population maintenance. The first one is concerned with mating selection and variation (e.g., crossover and mutation) where we want to produce promising offspring, hopefully better than their parents. The other is concerned with environmental selection (aka elitism) where we want to avoid losing the very best solutions found during the search course. The second issue can also be generalised as archiving, a process of taking new solutions, comparing them with the old ones and deciding how to update the population/archive.
Unfortunately, over 99% current archiving do not have theoretical quality guarantee, with their archive set suffering from degeneration. In other words, spending more computational effort does not necessarily result in quality improvements. The only a few archiving methods that theoretically ensure quality, however, are of little practical value – they fail to maintain a representative set on diverse problems. The proposed research is to tackle this issue. Its aim is to bridge theory and practice for multiobjective archiving, via exploring how archiving can be provided theoretical quality guarantee while applicable in various practical situations. The project will be supervised by Dr Miqing Li and Prof Xin Yao.
Eligibility: First or Upper Second Class Honours undergraduate degree and/or postgraduate degree with Distinction (or an international equivalent). We also consider applicants from diverse backgrounds that has provided them with equally rich relevant experience and knowledge. Full-time and part-time study modes are available.
We welcome applications from highly motivated prospective students with a background in Computer Science or Operational Research with an interest in evolutionary algorithms or population-based randomised algorithms. A familiarity with multi-objective optimisation is desirable but not essential.
If your first language is not English and you have not studied in an English-speaking country, you will have to provide an English language qualification.
We want our PhD student cohorts to reflect our diverse society. UoB is therefore committed to widening the diversity of our PhD student cohorts. UoB studentships are open to all and we particularly welcome applications from under-represented groups, including, but not limited to BAME, disabled and neuro-diverse candidates. We also welcome applications for part-time study.