Also, in the last few months or so:
_Chance_ had a nice discussion of fuzzy logic, for the benefit of
statisticians. The author pointed out that there was essentially
nothing that has *SO FAR* been achieved by fuzzy logic -
mathematically speaking - that could not be obtained much more
directly by conventional methods, e.g., regression, .... Where
Fuzzy Logic has succeeded brilliantly is in salesmanship, i.e.,
self-promotion.
- If there is more than that going on, I will be interested to
hear of it.
The article pointed out that engineers are frequently given abysmal
educations in statistical matters, so they do not recognize that
there HAS to be some intelligible model inside of the Black Box.
Also, I recommend Michael Smithson's _Ignorance & Uncertainty_
(Books in Print alphabetized under "_Ignorance &" , using ampersand,
PLUS it mis-spelled the author's name - but it is a fairly recent
book, so it still should exist).
Rich Ulrich, biostatistician wpilib+@pitt.edu
Western Psychiatric Inst. and Clinic Univ. of Pittsburgh