- author = "Marc-Andre Gardner",
- title = "Controle de la croissance de la taille des individus en programmation genetique",
- school = "Universite Laval",
- year = "2014",
- address = "Quebec, Canada",
- keywords = "genetic algorithms, genetic programming, bloat",
-
URL = "
http://hdl.handle.net/20.500.11794/25386",
-
URL = "https://corpus.ulaval.ca/jspui/handle/20.500.11794/25386",
-
URL = "
http://www.theses.ulaval.ca/2014/31215/31215.pdf",
- size = "167 pages",
- abstract = "Genetic programming is a hyperheuristic optimization approach that has been applied to a wide range of problems involving symbolic representations or complex data structures. However, the method can be severely hindered by the increased computational resources required and premature convergence caused by uncontrolled code growth. We introduce HARM-GP, a novel operator equalization approach that adaptively shapes the genotype size distribution of individuals in order to effectively control code growth. Its probabilistic nature minimizes the overhead on the evolutionary process while its generic formulation allows this approach to remain independent of the problem and genetic operators used. Comparative results are provided over twelve problems with different dynamics, and over nine other algorithms taken from the literature. They show that HARM-GP is excellent at controlling code growth while maintaining good overall performances. Results also demonstrate the effectiveness of HARM-GP at limiting overtraining and overfitting in real-world supervised learning problems.",
- resume = "La programmation genetique (GP) est une hyperheuristique d'optimisation ayant ete appliquee avec succes a un large eventail de problemes. Cependant, son interet est souvent considerablement diminue du fait de son utilisation elevee en ressources de calcul et de sa convergence laborieuse. Ces problemes sont causes par une croissance immoderee de la taille des solutions et par l'apparition de structures inutiles dans celles-ci. Dans ce memoire, nous presentons HARM-GP, une nouvelle approche resolvant en grande partie ces problemes en permettant une adaptation dynamique de la distribution des tailles des solutions, tout en minimisant l'effort de calcul requis. Les performances de HARM-GP ont ete testees sur un ensemble de douze problemes et comparees avec celles de neuf techniques issues de la litterature. Les resultats montrent que HARM-GP excelle au controle de la croissance des arbres et du surapprentissage, tout en maintenant de bonnes performances sur les autres aspects",
-
notes = "In French
supervisor: Christian Gagne",
