Re: Opinions on Kosko?

Mark A. Scarton (mascarton@wpmail.code3.com)
Mon, 10 Jun 1996 11:47:42 +0200


One of the most interesting aspects of Kosko's writings for me is his ability to
examine issues from multiple points of view. In particular, I've found his
insights into fuzzy sethood from a geometric modelling perspective to be
particularly insightful. Rather than a single (complex) mathematical proof, he
will also provide intuitive/conceptual examinations along several other lines of
reasoning.

In this, his work more resembles the Rennaisance inventors. He discusses an
issue from multiple perspectives, ensuring that his conjectures make sense not
only in the abstract but in a synergistic perspective with other working
principles and real world phenomena. That may not please the average
mathematician in academia, but it tickles the heck out of a practicing engineer.

>From a personal perspective, I've been able to use the principles put forth in
his articles, books, and dissertation in military systems (principally signal
processing, pattern recognition and matching, and correlations) and in medical
information systems (expert systems, information measures). They not only work,
but they work well, are fairly easy to implement, and can be validated for
correctness (try that with many neural nets!).

I can't really speak to Kosko's tendency to argue philosophy and east vs. west,
since I'm not trained in the area (I'm only an engineer). But I certainly share
his feelings/distain for the tendency of many "scientists" to constantly work in
the abstract and to ignore the impingements of the real world. I've seen too
many systems blow up, killing and maiming people in the process, due to the
failure to account for systemic complexity and the interactions found in the
real world. (Variable independence ... hah!)

Anyway, my $0.02US. I'll jump down off of my soapbox now. ;-)

Mark A. Scarton, ABD
CompUtah!, Park City, Utah USA
Home: 801.565.9835
Office: 801.265-4612