Newsgroups: comp.ai.genetic
Path: cantaloupe.srv.cs.cmu.edu!rochester!udel!news.mathworks.com!europa.chnt.gtegsc.com!news.sprintlink.net!demon!uknet!gdt!uwe-bristol!pa-ross
From: pa-ross@pat.uwe.ac.uk (PA Ross)
Subject: Assignment
Message-ID: <1995May11.120452.7303@pat.uwe.ac.uk>
Sender: usenet@pat.uwe.ac.uk (uwe nntp usenet poster)
Nntp-Posting-Host: usenet@pat.uwe.ac.uk (uwe nntp usenet poster)
Organization: University of the West of England, Bristol
Date: Thu, 11 May 1995 12:04:52 GMT
Lines: 39

Dear All,

My class has been give a genetic algorithm assignment to do, and I understand
and have written most of it. However there is one point I don't understand.
The main algorithm is :

Create Initial Population
DO
	Evaluate Indivdual Fitness
	Select Parent Individuals for recombination
	Perform recombination of parents
	Replace weak population members
UNTIL Some stopping Criteria

All is clear but 2 points.

Firstly I have been given two methods of creating offspring, crossover and
mutation. I have also been given rates for these, 0.8 + 0.01 but I don't
understand what these numbers mean ? If crossover is performed 80% if the
time and mutation 1% of the time, then what method do I use for the remaining
19% ?

Also during the slection we have been told to impliment an algorithm that
creates an probability of combining depending on there fitness at solving a
problem. Should I therefore calculate the indivuals probability and then
combine the top 20% ? Or should I calculate the probability of each person
reproducing then use random numbers to see if the person is successful and
creating an off spring ?

Thanks,

Paul

PS I'm also interested in finding some documentation that deals with writng
	memory resident interupt handlers. Any pointers would be most
	appreciated.



