3 ms·
Very basic but also very instructional. The key to these sort of algorithms is to write a very efficient fitness function. The best GAs use GA to fetch the best
by jordhy 15y ago
Very basic but also very instructional. The key to these sort of algorithms is to write a very efficient fitness function. The best GAs use GA to fetch the best fitness function (sort of like meta programming or "meta-reflection").
This PPT has a great intro on GAs: http://cs.uga.edu/~khaled/ECcourse/chapter03.ppt http://cs.uga.edu/~khaled/ECcourse/chapter03.ppt
- sukuriant 15y agoIt's so much more complex and amazing than that, especially in problem spaces that have numerous local optima. My personal favorite weird genetic algorithm exploration paradigms are where you start looking at the behavior of the genome; and, instead of searching out best behavior, you search out ~different~ behavior. It's called Novelty Search, and it's awesome!