Expert systems :: a sudoku expert — OZONE Asylum, home of the Mad Scientists

This page's ID: 27810

Edit Post

Who can edit a post?	The poster and administrators may edit a post. The poster can only edit it for a short while after the initial post.

Your User Name:
Your Password:
Login Options:	Remember Me On This Computer

Your Text: Insert Slimies » Insert UBB Code » Close Last Tag \| All Tags UBB Help	If I remember correctly, you need to keep some of the best samples from the population intact, while some are mated. Decent samples are also mated, while the worst samples are left to die. A few are mutated, and not mated but kept for comparison with the next population. But that's from a GA adapting a Neural Network which controlled a "minesweeper" robot. The sweeper had two "motors" which controlled a steering similar to tank tracks, as output. The sensors included distance and heading to the closest mine, distance and heading to the closest sweeper(s), a counter for successful sweeps and something more I think. The genetic code was a string of bits which somehow told the motors which speed they should run at, depending on the sensor values. (Think they acted like weights or something like that, I read the article a long time ago) Over time, the most successful sweepers got better and better "search patterns". They were able to clear more and more mines, which were respawned at a random position each time they got swept, during the evaluation run of the current generation. What would you count as the genetic code when mating and mutating? /TwoD Loading...
Options:	Enable Slimies Enable Linkwords

Edit Post