From: earlcox@earlcoxreports.com   
      
   In the simplest terms for those of us who actually build and deliver fuzzy   
   models,   
   variables are decomposed into their fuzzy sets (the "term set"). Thus   
   a variable like Height has a domain, say [3,10] (uom=feet), and Height   
   can be decomposed into Short, Medium, and Tall fuzzy sets (which,   
   of course, overlap). The variable Speed could be decomposed into   
   slow, fast, etc.   
      
   You cannot normally use Height and Short (as an example) in the same way   
   since they are different kinds of "Things". I can say,   
      
   if Height is Tall then weight is Heavy   
      
   or I can say   
      
   if Height is above around 8 then ....   
      
   (which converts (around 8) to a fuzzy number and then applies the above   
   hedge to   
   generate an S-shaped membership function over the expectancy width of the   
   number).   
      
   or I can say,   
      
   if height > 5 then ...   
      
   which is a crisp evaluation of the rule predicate  -- or in some fuzzy   
   models,   
   such as the one I use, the rule compiler would convert this to the   
   semantically   
   equivalent (above(near(5))   
      
   To  get hung up on the abstruse and technical definition for a linguistic   
   variable   
   (the epistemology of the grammar over which the LV is defined and over which   
   is can take on valid values) is a generally meaningless exercise in the real   
   world.   
      
   Just an observation.   
   Earl   
      
      
      
      
   "Dmitry A. Kazakov"  wrote in message   
   news:rsq4ovo9sdb1p5lvftenbriumbu3njfpip@4ax.com...   
   > On Tue, 07 Oct 2003 08:02:56 +0530, scooter    
   > wrote:   
   >   
   > >Just to clarify, are you saying that HEIGHT, SHORT, MEDIUM and TALL   
   > >are all FLVs?   
   >   
   > HEIGHT is a fuzzy set, so it can be viewed as a FLV.   
   >   
   > >If so, in the lingo, how do you differentiate between HEIGHT and its   
   > >subsets? (so I know what to call my Java classes)   
   >   
   > I do not understand the problem. A set is always different from the   
   > set of its subsets.   
   >   
   > You cannot mix HEIGHT and, say, SHORT in one context because they are   
   > defined over different domains. If you want to mix them, you have two   
   > choices:   
   >   
   > 1. You stay in the domain {SHORT, MEDIUM, TALL}. Then you can mix   
   > HEIGHT and {SHORT} (note the braces around SHORT).   
   >   
   > 2. You go to the original domain (R+). For this you "accumulate"   
   > HEIGHT to get a corresponding fuzzy set in the original domain (R+).   
   > This accumulated HEIGHT can be then mixed with SHORT.   
   >   
   > ---   
   > Regards,   
   > Dmitry Kazakov   
   > www.dmitry-kazakov.de   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)