From: earlcox@earlcoxreports.com   
      
   I disagree completely with this.The parallelism implicit in the evidence   
   gathering and composition management facilities of a true fuzzy system   
   provide an extremely high level of both meta-data and meta-knowledge   
   abstraction. Further, without such parallelism, it is impossible to run all   
   the rules that complete the topology (actually the morphology) of the   
   outcome fuzzy sets. I agree that hybrid rules incorporating fuzzy sets and   
   fuzzy propositions in traditional expert systems are extremely useful, but,   
   again, the disassociation between the degree of truth in an antecedent and   
   the degree of truth in the consequent in such systems reduces the reasoning   
   algebra to a form of interval arithmetic. But such a hybrid approach -- and   
   I have used such an approach a few times -- does allow you to write rules   
   such as the ones I specified in my previous commentary.   
      
   Your comment about stability also raises an issue that I see on this new   
   group occasionally. After nearly thirty years of writing and delivering   
   commercial fuzzy systems in many capital and time-critical applications I   
   have never had a case of the system suddenly becoming unstable. Naturally   
   I've seen the system do the wrong thing because of a design error or a   
   programming bug. These are not, in my mind, instances of instability. In all   
   the systems that myself, Bill Siler, and others have written and delivered   
   and are running daily against high transaction volume, high data volume, and   
   are evaluating multiple states in near-real time I have never been involved   
   with issue so instability. Performance tuning, yes. Feature extensions or   
   modification, yes. Bug fixes, yes. Stability, no. The Sendai ATO train   
   operator in Japan carries over two million people a day using a fuzzy   
   controller and so far hasn't gone berserk and crashed into noon express to   
   downtown Tokyo.   
      
   So my question is simply this -- where are all these unstable fuzzy systems   
   that generate so much concern about stability in a fuzzy system?   
      
   Earl   
      
      
   "Guillaume"  wrote in message   
   news:3fff4c1f$0$6966$7a628cd7@news.club-internet.fr...   
   > > Not necessarily. You must remember that a fuzzy system is, in effect, a   
   > > parallel processing system. All the rules are essentially run in   
   parallel as   
   > (snip)   
   > >    if A is High and B is Low then C is Elevated;   
   > >   
   > >    if C is Elevated then D is Small;   
   >   
   > I believe that to be actually a big limitation with "ordinary" fuzzy   
   > systems. Sequential rules (as in your example) could prove very useful,   
   > in my opinion. That would allow to build systems with much higher an   
   > abstraction level. Whether it would not exhibit some nasty stability   
   > problems is an interesting project to work on.   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)