Re: Kosko Profile in IEEE Spectrum

Walter Raisanen (azi@crl.com)
Sat, 2 Mar 1996 00:07:49 +0100


Peter Hamer (pgh@bnr.co.uk) wrote:

[good argument removed to save space]

: Personally, I find it difficult to imagine any form of confirmation that
: does not involve asking people what they think. Which rather naturally leads
: to the value of the membership function at any point being the fraction of
: those asked who agreed with the proposition

This is true in the gedanken arguments, but I am concerned with the
application of fuzzy logic to electrical engineering design problems.
In those, the last step in the design analysis process inevitably
requires 'crispifying' the fuzzy output so it can be interpreted
by the hardware. In all the cases I have studied (not exhaustive of course)
the final result could have been obtained faster by various
piece-wise linear appro9ximation methods.

[snip]

: Personally I would like to see more comparisons of the performance achievable
: by the two approaches. This may well require the different approaches to be
: applied by those familiar with them. [I'm very bored by demonstrations that
: method X (eg neural nets) outperforms linear regression -- on problems where
: the only excuse for applying linear regression can be ignorance of non-linear
: statistical methods.]
Me too! Lets leave the realm of belief systems and get back to science
and engineering.

-- 
<Image>