First thing Mathematix: How are you changing the font size when you quote me?


If the mathematical foundation of ANNs was solid you would be able to show me (with a proof) exactly how many hidden layers and neurons are required for a given number of inputs and outputs to optimise the network. No one has come even close to doing this. QED. (There was a thread on comp.ai.neural-nets recently about mathematical foundations which i''m sure you could reference if you so wish)

I don''t have a problem with you suggesting more theory, I just disagree that my tutorial needs it. From the dozens of emails I am getting I *know* I have achieved what I set out to do, that is, to bring ANNs to a wider audience and to stimulate people to go out and research the subject further.

I''m sure some readers would appreciate you doing a section on the maths though, it''d be nice to have a link to it for people to quickly refer to if they are mathematically inclined. (although doing all those equations is not a pleasant experience in MS Word I can tell you!)

quote: Original post by fup
First thing Mathematix: How are you changing the font size when you quote me?

He''s just doing a "quote" when he replies. It automatically puts things into a smaller font.

As I said on my page, by the way....good tutorial, fup!




Ferretman

ferretman@gameai.com
www.gameai.com

From the High Mountains of Colorado

Ah! Thanks Steve, I didn''t see that little ''quote'' button hiding away in the corner.

quote: Original post by fup
First thing Mathematix: How are you changing the font size when you quote me?


If the mathematical foundation of ANNs was solid you would be able to show me (with a proof) exactly how many hidden layers and neurons are required for a given number of inputs and outputs to optimise the network. No one has come even close to doing this. QED. (There was a thread on comp.ai.neural-nets recently about mathematical foundations which i''m sure you could reference if you so wish)

I don''t have a problem with you suggesting more theory, I just disagree that my tutorial needs it. From the dozens of emails I am getting I *know* I have achieved what I set out to do, that is, to bring ANNs to a wider audience and to stimulate people to go out and research the subject further.

I''m sure some readers would appreciate you doing a section on the maths though, it''d be nice to have a link to it for people to quickly refer to if they are mathematically inclined. (although doing all those equations is not a pleasant experience in MS Word I can tell you!)

I SAID: "The practical theory behind neural nets IS solid, it''s the relationship between a network and its application where everything becomes fuzzy!"

AT THE BOUNDARY WHERE YOU TRY TO APPLY NEURAL NETS IS WHERE THE "CONFUSION" COMES IN. THE ACTUAL FACT THAT THEY CAN LEARN MEANS THAT THE THEORY BEHIND THE ACTUAL NEURAL NETWORK CONSTRUCTION MUST BE CORRECT - DAMMIT!

DAMMIT - DAMMIT - DOUBLE DAMMIT!!! DID I NOT SAY THAT I LIKE YOUR SITE??? TRIPLE DAMMIT!


Regards,
Mathematix.

hehe, I''ll not comment any further.

hehe, I''ll not comment any further.

quote: Original post by fup
hehe, I''ll not comment any further.

It''s cool! It''s cool! I''m cool.


Regards,
Mathematix.

i looked through it briefly, and i like it. now how about one for genetic programming? Please?

I do not have to destroy, because I do not fear.

Genetic programming is a little too esoteric for the general public so I won''t be covering it accept for maybe a brief mention of it. My next tutorial will *sort* of go down that path though methinks.

GP is an extremely interesting subject but to do it right you really need a LISP(or similar) compiler which most people do not have. (I don''t think there are any free ones out there but if you know of one please tell me).

quote: Original post by fup
Genetic programming is a little too esoteric for the general public so I won''t be covering it accept for maybe a brief mention of it. My next tutorial will *sort* of go down that path though methinks.

Why do you believe that genetic programming is too esoteric? As far as I know genetic programming doesn''t require that much math knowledge, compared to NNs.

As for your tutorial, I haven''t read it and I am therefore not able comment on it.



/Mankind gave birth to God.