Re: Neural Networks, Programming Languages

  • From: "black ares" <matematicianu2003@xxxxxxxxxxx>
  • To: <programmingblind@xxxxxxxxxxxxx>
  • Date: Thu, 16 Oct 2008 00:59:36 +0300

please, look at some examples in lisp.
More than that I recomand you to see some web semantic examples in lisp.
This is the best language to suport the learning feature.
and this learning feature is one of the most important in A.I.
----- Original Message ----- From: "Ricks Place" <OFBGMail@xxxxxxxxx>
To: <programmingblind@xxxxxxxxxxxxx>
Sent: Tuesday, October 14, 2008 11:52 PM
Subject: Re: Neural Networks, Programming Languages


Thanks Guys:
First, This project sounds like an extreme challange but you guys and Old WillPearson use to talk about this stuff while I was still trying to learn what a class, method and property were.So thanks for putting up with what must seem to be really basic questions in an area I am not likely qualified to even scratch the surface of.I get Perceptrons. The math surrounding the more complex applications sounds daunting including Matrix Algebra upto Chaos Theory etc ( I can't remember how to spell Gausian and even the names of a couple of other mathfields mentioned ). I made it through the Calc series but any math beyond there I would have to learn from e-books, or articles, and I'm not sure how possible that would be. The Perceptron articles I read used figures, pictures, for many examples and formulas and they did not have any examples for beginning programmers to code up. So, LISP is the way to go eh? Do you remember any good primers on building very beginner orientated AI Neural Network applications with programming examples to try out in any language?
Simple single layer forward only, oh darn, you know the right words...
Note, I do not mind buying a e-book if it's accessible.
I will research LISP on Google, do you have a favorite flavor, compiler or whatever development environment you use? I am still trying to figure out if I can learn enough AI to create a Neural Network tuned to do some Stock Market Analysis to augment my own Fundemental Analysis techniques.
Rick USA
Rick USA
From: "Sina Bahram" <sbahram@xxxxxxxxx>
To: <programmingblind@xxxxxxxxxxxxx>
Sent: Tuesday, October 14, 2008 3:51 PM
Subject: RE: Neural Networks, Programming Languages


Please email me off list:

sbahram@xxxxxxxxx

Take care,
Sina

-----Original Message-----
From: programmingblind-bounce@xxxxxxxxxxxxx
[mailto:programmingblind-bounce@xxxxxxxxxxxxx] On Behalf Of black ares
Sent: Tuesday, October 14, 2008 3:28 PM
To: programmingblind@xxxxxxxxxxxxx
Subject: Re: Neural Networks, Programming Languages

I am interested in evry kind of documentationon lisp so if you can ofer
something...
More than the language it self, lisp ofer a variety of dialects (scheme) or
other like it.
I remember that I don't know at all tcl/tk, but with my lisp knowledge I was

able to debug a program in tcl/tk.
So if you have any good book on lisp and you are so kind to share it with
me, i wait for it.
I use lisp in my programming job.
best regards
Black Ares
----- Original Message ----- From: "Sina Bahram" <sbahram@xxxxxxxxx>
To: <programmingblind@xxxxxxxxxxxxx>
Sent: Tuesday, October 14, 2008 10:04 PM
Subject: RE: Neural Networks, Programming Languages


To comment a little on this.

Lisp is not a functional programming language. It supports a myriad of
programming styles which is why it has survived and thrived in some
instances for over 50 years.

Lisp supports functional, procedural, logical, aspect oriented, and object
oriented styles just to name a very small few.

Also, it is truly a wonderful language. In less than 300 lines of code, I was able to write a distributed evolutionary algorithm to do the class of
traveling salesman like problems efficiently and quite nicely.

As another example, the entire prolog language was implmented in lisp in
the
back of my introduction to lisp book, for example. Some of the three and
five liners in lisp are truly mind blowing.

*chuckling*, and folks think I'm a java guy, *snicker*, man do I have them
fooled.

I wish I had an opportunity to use lisp on a daily basis.

Take care,
Sina

________________________________

From: programmingblind-bounce@xxxxxxxxxxxxx
[mailto:programmingblind-bounce@xxxxxxxxxxxxx] On Behalf Of black ares
Sent: Tuesday, October 14, 2008 2:56 PM
To: programmingblind@xxxxxxxxxxxxx
Subject: Re: Neural Networks, Programming Languages


You can make A.I. even with visual c#, but you must make a lot of code.
There are other programming paradigms more dedicated to this.
1. Functional programming (represented by lisp/scheme)
2. Logical programming (represented by prolog)

for example a gcd calculation in prolog will look like:
gcd(x,0,X).
gcd(X,Y,D):-R is X MOD Y,gcd(Y,R,D).

Thats all
now you can make
gcd(15,5,X).
and in d you will have the result.

Compare this with a traditional procedural/oop programming euclid
algorithm
and you will see the benefits.
For having A.I. is necessary to have the learning capability in your
software
that means that a method have to change is behavior from a calling to
another.
To be more explicit let say we have a method that receive 2 parameters
x and y to integers
at first call the method returns the sum of the two numbers.
Some where in time at another call the same method with same parameters
have
to return the product of the two integers.
that according to some lines which are between those two calls.
And this change in behavior must be not predictiv
so you can not do an if in the method and according to an conditon return
either the sum or the product.
For this situation functional programming respond better than other
paradigms.
Because in a functional programming language you can modify the body of a
function easy from call to call.
in c# in dotnet 2.0 this is posible only using reflection.
dotnet 3.0 introduced 2 new things that I like very much
1. linq
2. f#

Also you can find on the net libraries written for c# used to interpret
lisp/prolog.
I worked with some of them in time to get faster results.

best regards.
bBlack Ares

----- Original Message ----- From: Ricks Place <mailto:OFBGMail@xxxxxxxxx>
To: programmingblind@xxxxxxxxxxxxxx
<mailto:programmingblind@xxxxxxxxxxxxx>
Sent: Tuesday, October 14, 2008 1:40 PM
Subject: Neural Networks, Programming Languages

Hi Guys:
Just looking at overviews of how Neural Networks are used to predict
Stock Market movements. Is the AI done an any language like Vb.net or are
there specific Programming languages used here. I am just looking at the
possibilities to see if my programming and math knowledge would allow me
to
play with them some.
I have a good background in math, stats and Vb.net and can, of
course, learn other things if necessary but what might they be?
Rick USA


__________
View the list's information and change your settings at
//www.freelists.org/list/programmingblind



__________
View the list's information and change your settings at
//www.freelists.org/list/programmingblind

__________
View the list's information and change your settings at
//www.freelists.org/list/programmingblind



__________
View the list's information and change your settings at //www.freelists.org/list/programmingblind



__________
View the list's information and change your settings at //www.freelists.org/list/programmingblind

Other related posts: