About the influence of features of fitness-functions on the convergence of the genetic algorithm
The adaptive capabilities of a parallel version of a multipopulation genetic algorithm are considered depending on the characteristics of certain classes of fitness-functions. Ways are proposed to increase the rate of convergence to the optimal solution based on effective control of algorithm parame...
Збережено в:
| Дата: | 2020 |
|---|---|
| Автори: | , |
| Формат: | Стаття |
| Мова: | Russian |
| Опубліковано: |
PROBLEMS IN PROGRAMMING
2020
|
| Теми: | |
| Онлайн доступ: | https://pp.isofts.kiev.ua/index.php/ojs1/article/view/428 |
| Теги: |
Додати тег
Немає тегів, Будьте першим, хто поставить тег для цього запису!
|
| Назва журналу: | Problems in programming |
| Завантажити файл: |  |
Репозитарії
Problems in programming| Резюме: | The adaptive capabilities of a parallel version of a multipopulation genetic algorithm are considered depending on the characteristics of certain classes of fitness-functions. Ways are proposed to increase the rate of convergence to the optimal solution based on effective control of algorithm parameters and strategies for the exchange of chromosome-solutions between populations. The results of computer experiments with the optimization of fitness-functions with various ratios of insignificant and significant factors are presented. The dependence of the convergence rate of the algorithm in the presence of a random effect on the values of fitness-functions is studied.Problems in programming 2020; 2-3: 362-367 |
|---|