Robust Configuration of Evolutionary Algorithms

Funding Source: ARC Discovery Project

The purpose of this research is to develop an intelligent framework for the robust configuration of evolutionary algorithms. This research is driven by the fact that the current design of evolutionary algorithms is sub-optimal and ineffective for many problem domains. In the proposed framework, a configuration is evolved while the algorithm is running for problem solving to ensure robust design. Its scientific outcomes include a novel framework for the automated design of algorithms and new techniques for exploiting assumptions in algorithmic design that may have been overlooked. Its practical outcomes include the provision of a robust problem-solving tool, strong research training and high–impact publications.

Return to Trusted Autonomy >>