Computing strategy adaptation for genetic algorithms

The article presents the results of comparative tests of different modification calculation strategy genetic algorithms. It compared the 5 popular adaptation for 5 the features representing different classes of problems. The results in-dicate that various modifications in the different issues under consideration. These tests allow you to create a genetic strategy that will automatically adapt to the problem. Were compared with each other classic genetic algorithm, and such modifications as messy genetic algorithms, uneven mutation, elit-ism and floating-point gene representation.