Monday 29 Sep 2014: Hyper-heuristics: Improving the Heuristic Search Process via Learning
Dr. Ender Ozcan - School of Computer Science, University of Nottingham
Harrison 103 15:00-16:00
A hyper-heuristic is a high level methodology that controls or generates
low level heuristics (or heuristic components) for solving computationally hard problems. Hyper-heuristics are often structured as low cost algorithms which are general and can be reused on unseen problem instances as well as other problem domains, desirably without any domain expert intervention.
There is a growing interest towards the use of data science techniques as components of adaptive (meta/hyper-)heuristic approaches. This talk will provide a brief overview of hyper-heuristics and cover some recently proposed general-purpose hyper-heuristics embedding machine learning methods for improving the heuristic search process.