Hacker News new | past | comments | ask | show | jobs | submit login

I was recently working on simple demos for Genetic Algorithms that only use basic, programming techniques.

My first one just tries to make a 64-bit bitting that is all 1's. I made random selection, roulette wheel, and tournament. Random is there as empirical proof that GA's are smarter than chance. Tournament is doing a great job.

Next idea is student loans. For maybe 15 loans with different rates, what payment plan knocks them all out the fastest? I'm fixing the payment size with a payment plan being a list of numbers representing which loan to apply the payment to (or 0 for no payment). Each step is (a) apply payment, (b) advance all loans with interest, and (c) see if the total is zero. Fitness will be interesting given I need to optimize for minimal principle and payments.

I found some other easy examples people did online that beginners might want to recreate. They include trying to get from point A to B with bananas at random places (avoid falling), best combo of items in your backpack for travel, and ants finding the best path to get the most food in a 2D map. I'll try to dig up the links later if someone wants them. Otherwise, those descriptions alone might be a starting point.






Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: