Genetic Algorithms and Machine Learning

Hello /prog/.

My introduction to the idea of genetic algorithms for machine learning was recent, and I've read a pretty good deal about them, but I want to test my knowledge.

Any good ideas of a project I could do to try writing a genetic algorithm?

>>1
Determine the most efficient read pattern to load a highly fragmented file from disk into RAM.

>>2
Are you trying to make him crash his har-- ohh.

>>3
I actually thought it would make for an interesting project to deal with bad swap file allocation, as the results from a generalized study are (were?) important to optimizing disk IO subsystems for large fragmented files. You probably have a point, though.

Let's see... what about learning how to play Tetris? Make it try to find the best and most efficient series of key-presses to get a good score, given read access to the current piece, the next piece, and the current score.

To help you get your head around the very weird world of Tetris strategies, check this out: http://tetris.wikia.com/wiki/Playing_forever

NIGGER I INVENTED C, BITCH >>4 WHAT THE FUQ

>>4

Thanks. Seems interesting and doable. If I knew more about filesystems, >>2 seems interesting as well.

The Jews are after me

Write a Megaman bot.

The Jews are after me

>>9
The Jews are after him

>>10
The JEWS are after him

>>11
The JEWS are after me.