probability vs. fuzziness
WILLIAM H. WOODALL (WWOODALL@alston.cba.ua.edu)
Mon, 11 Mar 1996 12:51:41 +0100
Although I find the philosophical debate on this topic interesting,
it seems more important to consider the practical applications of
fuzzy logic from a probabilistic point of view. We have spent the
past four years or so studying fuzzy control methods, fuzzy quality
control charts, fuzzy regression, and fuzzy forecasting. Three of
our papers have appeared in FUZZY SETS AND SYSTEMS. All of our
results are summarized in the Technometrics paper I offered to mail
to those interested. In all cases we have found methods based on
probability which are equivalent to or better than the fuzzy
methods. We have yet to find a case where fuzzy logic can solve a
problem which cannot be solved using probability. In some cases,e.g.
control methods, the probabilistic approach has a much more
straightforward interpretation avoiding the useful, but ad hoc
mathematics of the fuzzy methods.
Since I take the probabilistic side and don't plan to do any more
research on this topic, I plan to remove my name from this list.
It has been a lot of fun!
Bill Woodall