fishwick@fish.cis.ufl.edu (Paul Fishwick) (03/23/90)
I think that Ted is right on track - it is good for people to question basic motives and results - this is what science is all about. I, too, have reservations about neural net competence. As far as interpolation of curves is concerned, we did some work that showed that a neural net performed better than linear approximations but much worse than approximation by series such as the Fourier series. This is not all that surprising once one converts the graphical neural net picture into its constituent equations, and notes the model being realized. Let's have more people like Ted and less people arbitrarily applying neural networks as a magical ointment to a given problem. Let's have more comparative research. -paul fishwick +------------------------------------------------------------------------+ | Prof. Paul A. Fishwick.... INTERNET: fishwick@bikini.cis.ufl.edu | | Dept. of Computer Science. UUCP: gatech!uflorida!fishwick | | Univ. of Florida.......... PHONE: (904)-392-1414 | | Bldg. CSE, Room 301....... FAX is available | | Gainesville, FL 32611..... | +------------------------------------------------------------------------+