BO> There are no weights associated with selections, ie. the player can NOT 
 BO> say that he wants to play in SA02 by 10 units, SA06 by 7 units and
 BO> so on...
More accurately, the weights associated with the selections are not
defined, but still ordered in 1st, 2nd, 3rd place and so on.
(just to show i'm actually reading your msg)
 BO> Restriction list is as follows:
Must you really do this in C/C++? i'm currently making a shift from
C/C++ to higher level languages, and found that lots of times, Tcl,
Perl, Make, and such, are better suited to the task. In your case, the
language of choice might be Prolog. Most programmers are only familiar
with less than five programming languages, often only two or even one;
often it takes like 400 lines of C to do a thing you can do in 40 lines
of Prolog or Perl, depending on the task.
Two or three years ago my favorite languages were Basic and 8086
intel-style assembly language (as opposed to unix-style syntax)...
 BO> I considered a NN solution similar to solution of TS problem with a 
 BO> Hoppfield network but could not really formulate it.
 BO> Note: I'm neither a computer engineering nor a mathematics student, so 
 BO> please try to keep your answer simple in terms of jargon.
You're using quite a bit of jargon yourself :-) (as far as obscure
abbreviations go)
Prolog may be a hard language in some sense, for that kind of problems
i think it would be more natural, ie easier to translate from human
language. I don't want to analyze all of this problem, but if Prolog is
not the answer, then a derivative of Prolog, enhanced for
optimizations, might be better.
I keep a copy of your message in case I change my mind (i.e. i might
get a urge to do math), although i'm no Prolog expert, and i'm not at
all aware of most of the math stuff you're talking about. At worse,
this message might be useless. :-)
matju
--- Terminate 4.00/Pro
---------------