If you''re interested in classifier systems then, in addition to Goldberg''s book, you should take a look at:

Holland, John H. Adaptation in Natural and Artificial Systems. Ann Arbor: University of Michigan Press (1975).

Holland, John H, Induction: Process of Inference, Learning and Discovery. Cambridge, MA: MIT Press (1986)

Also, check out the research conducted by Lashon Booker. He''s another giant of the field and has made some very interesting discoveries about the inner workings of classifier systems.

On a side note, I dug out my old research report I did when I wrote my classifier system, however it has been corrupted (bloody Microsoft). It''s still got all of the text but some of the other info needs redoing. I guess I could fix it up and post it here as an article, since it includes a good description of the workings of classifier systems and even has my code for the predator program!

Cheers,

Timkin

I haven''t gotten round to ordering the Goldberg book, but I''ve heard that it''s supposed to be a good source - I''ve got some books by Holland (the Adaptation book, for example) and Kauffmann (At home in the universe) that are pretty interesting.

The book that got me started on this stuff: Complexity, by Michael Waldrop - a very good read, and very inspiring!

Timkin >> I think that would be a Very interesting resource - I haven''t really found anything sufficiently tangible on the web to get me started properly. I''ve done some interesting stuff with GA, once I got an example program to show me the basic mechanics (that I later replaced entirely with my own code)

I don''t know exactly what I''d use CS for - maybe some game AI, to try it out - but I''d really like to get a grip on it

-----------------------
Always down - Never out

quote:

---

Original post by toblo
The book that got me started on this stuff: Complexity, by Michael Waldrop - a very good read, and very inspiring!

---

ROFL: That was also my first introduction to Holland''s work and one of the first books I read on complex systems and artificial life. A very nice book!

I''m starting a new job soon so I''m not sure how much time I''ll have to sort out the old research report... I''ll try and find a non-corrupt version and post it as is.

Cheers,

Timkin

Holy crapolini! That''s deep stuff!!!

*HEH* Mr Waldrop has a lot to answer for

Some other AI-fans I've talked to have started out the same way

I'd really like a look at your stuff - My math skills are nothing to talk about in public, but if I see an algorithm I can often get more out of it

As the programmers adage goes:
1) Read
2) Read code
3) Code



EDIT>> BTW: what domain did you use your CS in? I'm trying to think up some nice example domain to try stuff in - I was thinking about some pursuit/evasion game to let CS's train against eachother in

-----------------------
Always down - Never out

[edited by - toblo on October 31, 2002 4:35:48 AM]

quote:

---

Original post by toblo
I''d really like a look at your stuff - My math skills are nothing to talk about in public, but if I see an algorithm I can often get more out of it

---

I''ll see what I can do... things are rather hectic at the moment.. I have a paper due in a few days... and my best friend had a heart attack yesterday... but he''s alive, so that''s a start!

quote:

---

Original post by toblo
what domain did you use your CS in? I''m trying to think up some nice example domain to try stuff in - I was thinking about some pursuit/evasion game to let CS''s train against eachother in

---

Lol... I did a predator-prey simulation. This was chosen because I wrote the system as part of a fourth year project in biomathematics. I was trying to show that simple behaviours, like pursuit mechanisms, could be learned and based on very simple sensory information and a drive to eat. Needless to say the program was very successful and I did very well for the subject.

Cheers,

Timkin

-My best wishes for your friend''s health.. that must be a terrible experience :-/

About the test-domain, I''ve been thinking about one of the most impressive uses I''ve read about for GA - somebody who''d used it to evolve really good sequences of actions for a speceprobe for navigating through the solar system, getting to the target on time, and with small fuel expenditure.. maybe a CS could be used to control an enemy in some SpaceWar-like game ..?

- a nice weekend to you - hope your friend copes well.

-----------------------
Always down - Never out