How many dimensions in Cooptimization?. Jaśkowski, W., Krawiec, K., & et al , N. K. In GECCO 2011 Late breaking abstracts, pages 829--830, July, 2011. Association for Computing Machinery.
abstract   bibtex   
Co-optimization test-based problems is a class of tasks approached typically with coevolutionary algorithms. It was recently shown that such problems exhibit underlying objectives that form internal problem structure, which can be extracted and analyzed in order to drive the search or design better algorithms. The number of underlying objectives is the dimension of the problem, which is of great importance, since it may be a predictor of problem's difficulty. In this paper, we estimate the number of dimensions for Tic Tac Toe and the Density Classification Task.
@inproceedings{ jaskowski2011how,
  author    = {Wojciech Jaśkowski and Krzysztof Krawiec and Natalio Krasnogor et al},
  title     = {How many dimensions in Cooptimization?}, 
  abstract   = {Co-optimization test-based problems is a class of tasks approached typically with coevolutionary algorithms. It was recently shown that such problems exhibit underlying objectives that form internal problem structure, which can be extracted and analyzed in order to drive the search or design better algorithms. The number of underlying objectives is the dimension of the problem, which is of great importance, since it may be a predictor of problem's difficulty. In this paper, we estimate the number of dimensions for Tic Tac Toe and the Density Classification Task.},
  booktitle   = {GECCO 2011 Late breaking abstracts},
  isbn   = {978-1-4503-0690-4},
  month   = {July},
  organization   = {Association for Computing Machinery},
  pages   = {829--830} ,
  year   = {2011}
}

Downloads: 0