Hacker News Reader

Comment by aclissold | Hacker News Reader

(serious question) Do any genetic algorithms take something like this into account? Seems like it could be a good way to increase the odds of furthering successful generations.

View on HN

A lot of genetic algorithms already seed the next generation with copies of the top X fittest individuals, so it's essentially already done.

IIRC, it's common for GAs to have variability as the only source of variation, not crossbreeding. And those that do use crossbreeding usually don't have a concept of the "sex" of an individual algorithm. So, crossbreeding of what are effectively two haploid gametes (a polar body isn't exactly a gamete, but...) from the same diploid individual when mates of the opposite sex are in short supply isn't something that would be directly applicable to most genetic algorithms.

You could build a framework for evolving genetic algorithms where this would be relevant and meaningful, but I'm not sure there is any good reason to bother with all the complexity that would be necessary.

Gotcha. It's not that genetic algorithms are meant to emulate the real-world, but rather that they're inspired by it.

While possible, I imagine it would require carting around more baggage with each variation, and would also require you simulate scarcity. It's been a long time since I played with genetic algorithms a little in college, but I'm not sure they often model making it hard to find a mate.

Yeah, sure they could

Reply on news.ycombinator.com