Investigating the Guidance Feature of Searching in the Genetic Algorithm
IRAQI JOURNAL OF COMPUTERS, COMMUNICATIONS, CONTROL AND SYSTEMS ENGINEERING,
2014, Volume 14, Issue 1, Pages 21-34
AbstractAbstract: There is an argument about the optimization capability of the Genetic Algorithm (GA) and whether its approach of search is guided or random. Although it has its own criticism, Schema theorem was the main effort for explaining the way a GA works, and justifying that the GA search is guided. After that, Schema theorem efforts continued for analyzing and justifying GA optimization approach with no complete analytic proof so far.
In this paper a different approach, namely a statistical approach, is used to test and justify guidance feature in the Genetic Algorithm when it is used for optimization purposes. Different standard functions have been tried and different tests have been done, and the results obtained proved that the GA is a guided search method and not random at all. Moreover, the tests carried out and results obtained proved the importance and necessity of each of the operators or techniques used by the GA.
- Article View: 273
- PDF Download: 15