Re: Qestion!
- From: "comcast" <billyjoe385@xxxxxxxxxxx>
- Date: Mon, 2 Jan 2006 08:20:05 -0800
While having no formal training in AI, I can share my observations with you.
The probability of finding a good fit is low, therefore fast computers are
used. Also it is more probable to find a better fit by a small modification
to the algorithm of one already having a somewhat good fit than try a
completely random one.
As for uncertainty, I test the best ones on other similar problems to see
who handles the difference best.
But one thing is for certain, putting your homework on a news group is no
substitute for studding.
Bill
"armoos" <bocheh@xxxxxxxxx> wrote in message
news:1135707998.417875.213080@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
> 1) What is the impact of probability in genetic algorithms?
> 2) In genetic algorithms how do we model the uncertainty?
>
.
- Prev by Date: Re: Looking for information
- Next by Date: Proofs of Reinforcement Learning Algorithms
- Previous by thread: Re: Looking for information
- Next by thread: Proofs of Reinforcement Learning Algorithms
- Index(es):
Relevant Pages
|
