Newsgroups: comp.ai.fuzzy
Path: cantaloupe.srv.cs.cmu.edu!bb3.andrew.cmu.edu!newsfeed.pitt.edu!portc02.blue.aol.com!howland.erols.net!newsfeed.internetmci.com!in1.uu.net!uucp4.uu.net!sangam!konark!saathi.ncst.ernet.in!sasi
From: Sasikumar M <sasi@saathi.ncst.ernet.in>
Subject: Re: Defuzzification to Boolean??
In-Reply-To: <32BD81B2.17AE@bytecraft.com>
X-Nntp-Posting-Host: saathi.ncst.ernet.in
Content-Type: TEXT/PLAIN; charset=US-ASCII
Message-ID: <Pine.ULT.3.94.961224190829.6310A-100000@saathi.ncst.ernet.in>
Sender: news@konark.ncst.ernet.in (News Administration)
Organization: National Centre for Software Technology, Bombay, India
References: <32BAF12D.437C@mitre.org> <32BD81B2.17AE@bytecraft.com>
Mime-Version: 1.0
Date: Tue, 24 Dec 1996 13:39:46 GMT
Lines: 21


> Michael P. McLaughlin wrote:
> > 
> > I am a newbie (= 0.95).  I was wondering whether defuzzification to a
> 
> Yes it is quite common. There are a number of approaches that are used
> The one that I use most often is to declare a consequence with 
> two lingistic results "yes" and "no" and after defuzzification 
> compare the result to a treshold that is half way between the
> crisp value for yes and no. This will allow all of the fuzzy rules
> to be evaluated and accounted for and will still return an 
> boolean result.This has been used for a lot of applications 
> where we are turning on or off motors performing some binary
> action.


But is this any different from the certainty factor mechanism
of MYCIN-like systems? Basically what is FUZZY about this?

  - Sasi

