Improvement of functional and non-functional aspects software through search-based techniques

Investigate the use of search-based techniques to improve the functional and non-functional aspects of software. Search-based techniques are optimisation techniques guided by a fitness function, for example genetic algorithms, and stochastic hill climbing. "Functional aspects" meant the correct functioning of software, according to a test suite or a specification. The techniques researched would attempt to automatically fix bugs, using the test suite or a specification as a guide. Non-functional aspects include power consumption, execution time, memory use etc. The search-based techniques would act on the source code of the software system, aiming to make changes to improve it in the light of its functional/non-functional behaviour.