Evolving Distributed Algorithms with Genetic Programming: Election

Authors

Abstract

In this paper, we present a detailed analysis of the application of Genetic Programming to the evolution of distributed algorithms. This research field has many facets which make it especially difficult. These aspects are discussed and countermeasures are provided. Six different Genetic Programming approaches (SGP, eSGP, LGP, RBGP, eRBGP, and Fraglets) are applied to the election problem as case study utilizing these countermeasures. The results of the experiments are nalyzed statistically and discussed thoroughly.

Keywords

Election, Rule-based Genetic Programming, RBGP, Standard Genetic Programming, SGP, Fraglets, Linear Genetic Programming, LGP, Epistasis, Neutrality, Communication, Loops

BibTeX

@inproceedings{WZ2009EDAWGPE,
  author                    = {Thomas Weise and Michael Zapf},
  title                     = {{Evolving Distributed Algorithms with Genetic Programming: Election}},
  booktitle                 = {Proceedings of the First ACM/SIGEVO Summit on Genetic and Evolutionary Computation (GEC'09)},
  editor                    = {Lihong Xu and Erik D. Goodman and Yongsheng Ding},
  publisher                 = {{ACM Press: {New York, NY, USA}}},
  pages                     = {577--584},
  year                      = {2009},
  location                  = {{Hua-Ting Hotel {\&} Towers: {Sh{\`{a}}ngh{\v{a}}i, China}}},
  url                       = {http://www.it-weise.de/documents/files/WZ2009EDAWGPE.pdf},
  doi                       = {10.1145/1543834.1543913},
  key                       = {WZ2009EDAWGPE},
},

Links