home bbs files messages ]

Forums before death by AOL, social media and spammers... "We can't have nice things"

   comp.ai.fuzzy      Fuzzy logic... all warm and fuzzy-like      1,275 messages   

[   << oldest   |   < older   |   list   |   newer >   |   newest >>   ]

   Message 789 of 1,275   
   Bruce Postlethwaite to All   
   Re: What membership function to choose ?   
   25 Sep 08 09:51:24   
   
   e8f5b4fb   
   From: no_spam_b.postlethwaite@strath.ac.uk   
      
   It really doesn't matter much, it's what you do with them that's important.   
   You can construct a rule-base using any shape of membership function, the   
   rules and placements may end up slightly different, but if you have some   
   sort of optimising design method you'll usually end up with a reasonable   
   fit. In some applications, a particular shape of membership function may   
   give better results after optimisation, but it's impossible to tell a   
   priori.   
      
   Some identification methods do, however, rely on particular shapes. The   
   stuff we did at Strathclyde required triangular membership functions - using   
   these allowed us to reduce the identification problem to one that could be   
   solved using linear least-squares.   
      
   Bruce Postlethwaite   
      
   "Ray"  wrote in message   
   news:1275bbc1-da8a-42e8-8ec6-4d81b867191f@l42g2000hsc.googlegroups.com...   
   > Membership functions may have different shapes (triangular, bell-   
   > shape, trapezoid, etc.). Can anyone tell me please if there is any   
   > established criteria or rule of thumb that can help make a sensible   
   > choice of the membership function ? Thank you so much for your help.   
   > Ray.   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)   

[   << oldest   |   < older   |   list   |   newer >   |   newest >>   ]


(c) 1994,  bbs@darkrealms.ca