Genetic Algorithms in 3 Minutes
By ai-depot | June 30, 2002
What Next?
Facing Facts
Genetic Algorithms can search for good solutions quite well. They don’t mind dealing with a huge amount of possible options, also known as large search spaces. However, when the solutions are complex and the fitness function is non-continuous, GA tend to struggle a bit more.
No matter what techniques are used in the evolution process, representation always has a huge impact on the quality of the final solution and the time required to achieve that solution.
Almost all the solutions found by Genetic Algorithms are not quite optimal, this is called sub-optimality. This is due to the way solutions are combined by the genetic algorithm, which is ill-suited to local optimisations.
Future Work
Genetic Algorithms can search for good solutions quite well. They don’t mind dealing with a huge amount of possible options, also known as large search spaces. However, when the solutions are complex and the fitness function is non-continuous, GA tend to struggle a bit more.
No matter what techniques are used in the evolution process, representation always has a huge impact on the quality of the final solution and the time required to achieve that solution.
Almost all the solutions found by Genetic Algorithms are not quite optimal, this is called sub-optimality. This is due to the way solutions are combined by the genetic algorithm, which is ill-suited to local optimisations.
Written by Alex J. Champandard.
Tags: none
Category: tutorial |
