I’m re-re-re-re-starting trying to build a program that uses a genetic algorithm to draw images using triangles. Like my blog that I keep “restarting.” Like everything in my pathetic life. Last week I read An Introduction to Genetic Algorithms by Melanie Mitchell as a primer. It’s a good book that talks about the history, various technics and mathematics behind understanding GAs. That said, I could never shake the feeling that the central theme of the book was “They’re an interesting way to program things, but we’re not sure what they’re good for and are most likely just a bad form a hill climbing.” I wasn’t too bothered by this, because my goal with this program is to learn more about GAs and not build scalable, efficient machine learning algorithms.
I had an old version of this program that made pretty good images (see the panda on this earlier post, or this one). I then remade a shitty version of the program, like 2 years ago, that I managed to get working again as well. It takes longer and produces worse results, more proof that I am declining in my old age.
I sketched out a rough plan of my revised algorithm over the weekend, and now it’s finally time to implement. Too bad it’s been ages since I’ve written any code let alone ceepeepee. Wish me luck.