Newsgroups: comp.ai.neural-nets,comp.ai.fuzzy
Path: cantaloupe.srv.cs.cmu.edu!das-news.harvard.edu!news2.near.net!yale!yale!yale.edu!spool.mu.edu!howland.reston.ans.net!gatech!concert!sas!mozart.unx.sas.com!saswss
From: saswss@hotellng.unx.sas.com (Warren Sarle)
Subject: Re: Fuzzy Cognitive Map Question ?
Originator: saswss@hotellng.unx.sas.com
Sender: news@unx.sas.com (Noter of Newsworthy Events)
Message-ID: <CvwG9G.GI8@unx.sas.com>
Date: Sat, 10 Sep 1994 05:51:16 GMT
References:  <1702CF3A4S86.RVANRAAM@bcsc02.gov.bc.ca>
Nntp-Posting-Host: hotellng.unx.sas.com
Organization: SAS Institute Inc.
Followup-To: comp.ai.fuzzy
Lines: 59
Xref: glinda.oz.cs.cmu.edu comp.ai.neural-nets:18823 comp.ai.fuzzy:2865


In article <1702CF3A4S86.RVANRAAM@bcsc02.gov.bc.ca>, RVANRAAM@bcsc02.gov.bc.ca writes:
|> Fuzzy Cognitive Map Questions
|>
|> I am almost getting the hang of the Fuzzy Cognitive Maps
|> described by Dr. Rod Taber and Dr. Bart Kosko in the book
|> Fuzzy Thinking. I get some Fuzzy model described in the book
|> to work out like the African model but others like the
|> drug model still have some problems.

Fuzzy questions should be posted to comp.ai.fuzzy, not comp.ai.neural-nets.
I set the follow-up accordingly.

Fuzzy cognitive maps are ludicrously simplistic models for these
applications. If you want to do some serious modeling, take a look
at these books:

Bollen, K.A. (1989) Structural Equations with Latent Variables.
Wiley: NY.

Loehlin, J.C. (1987) Latent Variable Models. Lawrence Elbaum Associates:
Hillsdale, NJ.  (A 2nd edition is out now.)

|> For example:
|> 1) When I iterate by multiplying the state vector bu the
|>    state transition matrix the numbers get quite large so
|>  
|> a) Do I scale them by dividing every number in the matrix
|>    by the ABS of the largest matrix element ?  This works for
|>    some models.
|>  
|> b) Also when the value of a state becomes a negative value
|>    do I turn that state off by setting it to zero ?
|>  
|> c) Do the numerical values of the states mean anything or
|>    do they just indicate that things went up or down ?
|>  
|> d) How many iterations does it take to stabalize the drug
|>    model ?  Since the author was doing it in his head it
|>    doesn't seem like it took many.  My model doesn't seem
|>    to converge too fast (too fast = fuzzy term).
|>  
|> e) Does anyone have some code or is willing to check my code ?
|>  
|> f) Is this the right group for asking these kinds of questions ?
|>  
|> Thanks anyone.
|>  
|>   * * *
|>   Regards,
|>   Ray Van Raamsdonk  (389-3725)
|>   BC Systems


-- 

Warren S. Sarle       SAS Institute Inc.   The opinions expressed here
saswss@unx.sas.com    SAS Campus Drive     are mine and not necessarily
(919) 677-8000        Cary, NC 27513, USA  those of SAS Institute.
