Koskos book on Fuzzy Logic and Neural Networks should help
>Surely all, non-boolean network strategies, are "fuzzy", or does it
>depend on how the resultant outputs are processed.
Yes and there is argument for saying that all boolean networks are simply
special cases of fuzzy (ie. fuzzy but using rectangular membership
functions). In control this gets a bit confusing since over a small
signal range (small signal theory) all controllers can be represented by
PID therefore all fuzzy is a special case of PID. Its just that fuzzy is
a good/simple/easy to visualize method of doing it all.
J.D
------------------------------