1. [gen] a very simple first attempt atwriting an AI
Return-Path: <johnmcadam at clix.pt>
2001 18:57:05 -0000
2001 18:57:05 -0000
(195.23.132.13)
by inmta008.topica.com with SMTP; 14 Feb 2001
18:57:05 -0000
2001 18:49:58 -0000
by furao2.ip.pt with SMTP; 14 Feb 2001 18:49:58
-0000
2001 18:49:56 -0000
(195.23.205.9)
by perca.ip.pt with SMTP; 14 Feb 2001 18:49:56
-0000
Mail
id <01C096B7.9E8E95A0 at 195-23-205-9.nr.ip.pt>; Wed,
14 Feb 2001 18:54:54 -0000
=_NextPart_000_01C096B7.9EB94F20"
------ =_NextPart_000_01C096B7.9EB94F20
Hi, well I got the thing simplified about as much as
I can. Here is a
very small (180 lines) program to test 'Patrec', my
AI pattern recognizer.
It is probably too simple to be of much use, but I
think it has some value
at least for those who need a little introduction to
help get started on
their own AI projects. It is just a first draft of
course, that's why I hope to
get some feedback (I guess I'm like Patrec in that
respect) on what
I need to work on. I hope to be able to produce
Patrec as an include
file with some easy interface so that other programs
can just plug
Patrec in and let him (oops, IT 
) start figuring
things out. If somebody
comes up with some other tests please let me know.
Right now Patrec
needs a sequence holding the relevant information, a
set of questions
(boolean) about the relationships in the
information, a list of possible
responses and a scoring or feedback mechanism to let
her know if she
is doing well or not. Please let me know what you
think. When I get a
better, well tested version I'll post it to the RDS
contibutions page.
Run the program, hit enter a few times and watch the
%correct indicator
at the bottom of the page.
JOHN
2. Re: [gen] a very simple first attempt atwriting an AI
On 14 Feb 2001, at 11:46, John McAdam wrote:
>
> Hi, well I got the thing simplified about as much as
> I can. Here is a
> very small (180 lines) program to test 'Patrec', my
> AI pattern recognizer.
<blink> Where is it?
Kat
