Welcome to CHeSC 2011, the first
Cross-domain Heuristic Search Challenge

Citing HyFlex

If you use HyFlex for your research, we kindly ask you to acknowledge it in the respective publications by citing the following article:
HyFlex: A Benchmark Framework for Cross-domain Heuristic Search. G. Ochoa,  M. Hyde,  T. Curtois,  J. A. Vazquez-Rodriguez,  J. Walker,  M. Gendreau,  G. Kendall, B. McCollum, A. J. Parkes,  S. Petrovic,  E. K. Burke (2012).  European Conference on Evolutionary Computation in Combinatorial Optimisation (EvoCOP 2012), J.-K. Hao and M. Middendorf (Eds.), LNCS 7245, pp. 136-147. Springer, Heidelberg [bib entry].

HyFlex and Hyper-heuristic Resources

A page that collects bibliographies and resources is now available. Please send your contributions to gabriela.ochoa[at]nottingahm.ac.uk

CHeSC 2011 Benchmark

Compare your Hyper-Heuristic with the Competition Algorithms! The competition results now act as a benchmark of generality. Is your hyper-heuristic superior to the winners? Calculate your results with this program: chesc_benchmark.zip

Conference Special Session

Cross-domain Heuristic Search, Learning and Intelligent OptimizatioN (LION 6), Jan 16-20, 2012, Paris, France.

The first Cross-domain Heuristic Search Challenge seeks to bring together practitioners from operational research, computer science and artificial intelligence who are interested in developing more generally applicable methodologies. The challenge is to design a high level search strategy that controls a set of problem specific low level heuristics. The set of low level heuristics will be different for each problem domain, but the high level strategy which controls the heuristics must remain the same. The current prize fund is £ 3,000  to be split between the first (£ 1,500), second (£ 1,000) and third (£ 500) place competitors.   The top three competitors will be waived from the registration fee, and the accommodation fee. 

In order to run the competing high-level search strategies (hyper-heuristics) across different domains, we propose using a common software interface (HyFlex) for dealing with different combinatorial optimisation problems.  HyFlex provides a set of routines for generating and evaluating solutions; and applying low-level heuristics, such as move or mutation operations, ruin and recreate heuristics, recombination operations and hill-climbers, on these solutions.

The competition is organised and run by the Automated Scheduling, Optimisation and Planning (ASAP) group at the University of Nottingham, Nottingham, UK with partners from Queen’s University, Belfast, UK, Cardiff University, UK
and the École Polytechnique, Montreal,Canada.

Last Updated: 15 February 2012, by Gabriela Ochoa

- organising partners -

University of Nottingham

                University Belfast


- sponsors -