Led by Professor Richard Everson, Dr Jonathan Fieldsend, Dr Ed Keedwell and Dr Yiming Ying

Finding the "best" solution to a problem with a vast number of possibilities is well known to very difficult. In fact, many problems are NP-hard, precluding solution on even the fastest machines available.  The group at Exeter focuses on approximate methods, often inspired by natural systems for finding good solutions to hard optimisation problems.   Many optimisation problems have competing objectives, for example as performance is improved cost increases. Recognising this, much of the research in Exeter is on multi-objective (two, three or four competing objectives) and many-objective (more than four objectives) optimisation.


Improving air traffic safety:

Multi-objective optimisation of short term conflict alert systems that warn air-traffic controllers of potentially hazardous situations.

Multi-objective simulated annealing:

Simulated annealing, a well-known, provably convergent technique for optimising single objective functions extended to multi-objective optimisation.

Extra projects: info to follow

  • Multi-class ROC
  • PSO
  • Uncertainty
  • Visualisation of multi-objective optimisation