Abstract It is observed that there are two important issues in simulated search process of natural evolution: population diversity (exploration) and selective pressure (exploitation). These factors are strongly related: a strong selective pressure “supports” the premature convergence of the evolutionary search and a weak selective pressure can make the search ineffective. Thus it is paramount important to strike a balance between these two factors [19, 97]. Many of the various operators that are used to tune evolutionary search are really indirect means of affecting selective pressure and population diversity. As selective pressure is increased, the search focuses on the best individuals in the population, and when they mate, they usually produce almost similar offspring. Due to this increased exploitation, the population diversity is lost in the subsequent generations. Reducing the selective pressure or using larger population increases exploration because more individuals are involved in the search but the computational effort is increased thereby. Hence, most of the evolutionary algorithms are used for moderately small population size and this trade-off is pronounced due to mating of similar individuals.
An Incest Prevented Evolution Strategy Algorithm
2004-01-01
11 pages
Article/Chapter (Book)
Electronic Resource
English
Can aircraft crash fires be prevented
Engineering Index Backfile | 1947
|Rail corrugation -- Can it be prevented ?
Engineering Index Backfile | 1955
|