Pending GCC-3.1 upgrade - do or don't

Spencer Collyer spencer at lasermount.uklinux.net
Thu May 9 11:15:14 PDT 2002


On Wed, 8 May 2002 20:21:18 +0100, Ian Molton wrote:

> > Nope, sorry, have to correct you on this one. There have been a number
> > of studies using artificial systems (because they are a _lot_ easier
> > to study) where it's been demonstrated that the majority of the
> > improvement in evolutionary systems is caused by cross-over effects
> 
> Hmm. If I ever get subscribed to LFS-chat I would take this there, but
> you are wrong.
>
> evolution happens through natural selection. that is the process wherby
> things with a disadvantage die, and things with advantages live.
> 
> the method the things reproduce by is irrelevant to that.
> 
> the only difference the reproductive method might make would be to
> perhaps acclerate the mutation rate, but I dont have numbers for that.


> the point is that asexual reproduction does NOT result in a series of
> clones, which is effectively what the original poster suggested.

Actually, in almost all cases they are indeed clones of the original (in
the sense that their DNA is identical, which is the only real definition
that fits.)

If I had my copy of _The Blind Watchmaker_ readily to hand I'd grab it and
find the figures that Richard Dawkins gives for the mutation rate of DNA.

However, I do have to hand my copy of Goldberg's classic _Genetic
Algorithms in Search, Optimization & Machine Learning_, and on page 14 of
that book we find the following on the role of mutation:

"If reproduction according to fitness combined with crossover gives
genetic algorithms the bulk of their processing power, what then is the
purpose of the mutation operator? Not surprisingly, there is much
confusion about the role of mutation in genetics (both natural and
artificial). Perhaps it is the result of too many B movies detailing the
exploits of mutant eggplants that consume mass quantities of Tokyo and
Chicagom but whatever the cause of the confusion, we find that mutation
plays a decidedly secondary role in the operation of genetic algorithms.
Mutation is needed because, even though reproduction and crossover
effectively search and recombine extant notions, occasionally they may
become overzealous and lose some potentially useful genetic material (1's
or 0's at particular locations). In artificial genetic systems, the
mutation operator protects against such an irrecoverable loss. In the
simple GA, mutation is the occasional (with small probability) random
alteration of the value of a string position. In the binary coding of the
black box problem, this simply means changing a 1 to a 0 and vice versa.
By itself, mutation is a random walk through the string space. When used
sparingly with reproduction and crossover, it is an insurance policy
against premature loss of important notions."

To summarise then, it is the reproduction of the fittest entities by means
of crossover that leads to improvement, and mutation's role is to help
prevent loss of variety in the gene pool.

S>

-- 
<<< Eagles may soar, but weasels don't get sucked into jet engines >>>
7:46pm up 90 days, 4:12, 11 users, load average: 2.02, 1.71, 1.55
Registered Linux User #232457
-- 
Unsubscribe: send email to listar at linuxfromscratch.org
and put 'unsubscribe lfs-dev' in the subject header of the message



More information about the lfs-dev mailing list