Theorie von Schwarmverfahren und ihre Effektivität in unsicheren Umgebungen (TOSU)

• Antragsteller: Professor Dr. Tobias Friedrich, Ph.D.
• Fachliche Zuordnung: Theoretische Informatik
• Förderung: Förderung von 2014 bis 2017
• Projektkennung: Deutsche Forschungsgemeinschaft (DFG) - Projektnummer 247100267

• Erstellungsjahr: 2019

Zusammenfassung der Projektergebnisse

The theory of randomized search heuristics for uncertain environments has progressed tremendously over the past few years, in no small part due to the efforts of this project. A number of researchers of the field have picked up this topics and complemented or extended our results. This established core settings for the analysis of search heuristics in uncertain environments, shaping the research field. Overall, we now have a good feeling what kind of properties make a search heuristic excel in certain situations and stagnate in others. In particular, our research was not limited to swarm algorithms: Due to our generalization to EDAs we already have a much greater scope, but the insights gained point at interesting properties for crossover-based optimization, a topic notoriously difficult to analyze. Furthermore, we have set our results in perspective by giving analyzes for mutation-based optimization algorithms. These are at the core of the field, thus we contributed to the general understanding of randomized search heuristics in uncertain environments, not just to the subcategory of swarm algorithms.

Projektbezogene Publikationen (Auswahl)

• Robustness of ant colony optimization to noise. Genetic and Evolutionary Computation Conference (GECCO). Best Paper Award (ACO/SI Track). (2015), 17–24
  Friedrich, T., Kötzing, T., Krejca, M. S., Sutton, A. M.
  (Siehe online unter https://doi.org/10.1145/2739480.2754723)
• The benefit of recombination in noisy evolutionary search. International Symposium of Algorithms and Computation (ISAAC). Ed. by K. Elbassioni and K. Makino. (2015), 140–150
  Friedrich, T., Kötzing, T., Krejca, M. S., Sutton, A. M.
  (Siehe online unter https://doi.org/10.1007/978-3-662-48971-0_13)
• EDAs cannot be balanced and stable. Genetic and Evolutionary Computation Conference (GECCO). (2016), 1139–1146
  Friedrich, T., Kötzing, T., Krejca, M. S.
  (Siehe online unter https://doi.org/10.1145/2908812.2908895)
• The compact genetic algorithm is efficient under extreme gaussian noise. IEEE Transactions on Evolutionary Computation 21. (2017), 477–490
  Friedrich, T., Kötzing, T., Krejca, M. S., Sutton, A. M.
  (Siehe online unter https://doi.org/10.1109/TEVC.2016.2613739)