h2g2 The Hitchhiker's Guide to the Galaxy: Earth Edition

Genetic algorithms are great fun. I have never actually used the cross-over routine described; I've tended towards either random mutations or random crossovers, neither of which (now I come to think of it) are as effective. Still, I've managed to produce a program which writes C source code using a genetic algorithm to perform a fairly simple task (telling whether a letter of a standard typeface is uppercase or lowercase).

More interesting, in my opinion, is when you work out some way of letting the selection take care of itself. Then you're leaving the realm of the genetic algorithm and getting involved in artificial life... which has got to be a good thing. My best program involves little triangles (creatures) trying to catch little squares, with their movements being dictated by their genes. A successful creature creates copies of itself, meaning that - given time - they evolve...

It doesn't work *very* well, but it's fun to watch .

(As a completely irrelevant aside, I have in fact used 'Genetic Algorithm' as a name when playing various online computer games...)

26199

That was a nice and concise explanation .... shame I never found it a couple of years ago when I needed it.

The simplest soultion is the "random mutation path", but as the man said, not always the best....