genetic implementation
- From: Marco <cimmo@xxxxxxxxx>
- Date: Fri, 24 Feb 2006 17:19:07 +0100
Hi to all,
if I have a lot of rules like this:
object(name, strength, list of answers).
EXAMPLE:
object(apple, 100, [s(1, y), s(2, n)]).
object(apple, 150, [s(1, n), s(2, y)]).
object(apple, 200, [s(2, n), s(3, n)]).
object(apple, 250, [s(2, n), s(4, y)]).
and I want to make a genetic algorithm that produce new rules I have thought these steps:
- take all the rules with Strength >=X (X is fixed by me, for example 100)
- choose at random (rules with more strength have more probability to be choosen) a couple of rules that will produce two new rules
- choose a K that means that K answers will be exchanged between two rules
- make the mutation in a very small case in
is this a correct implementation in my case?
Thanx
Marco
.
- Prev by Date: Stipends available for MSc Intelligent Systems for EU students
- Next by Date: Brownian Tree
- Previous by thread: Stipends available for MSc Intelligent Systems for EU students
- Next by thread: Brownian Tree
- Index(es):
Relevant Pages
|
