fishwick@fish.cis.ufl.edu (Paul Fishwick) (04/25/89)
Myke Rynolds writes.. >Grossberg designed ART I in '76. BAM is nothing more than a solution to >systems of linear equations, which is equivolent to least means squared, >linear programming, linear regession, matrix psuedo-inverses, on and on... >You've nothing to dispute here, its what I ment in one of my earlier messages >when I refered to the deeper simplicity of math. Its an excedingly old trick >in a new form, cognition! >-- In addition to your suggestion that much of neural nets is really basic linear algebra, albeit couched in new terminology --- I would add that there appears to be not much difference between "training neural networks" and using adaptive control methods and methods of automatic system identification. Several structures have been used for automatic identification and neural nets seem to be one of many possible cases. Anyone is control theory/science out there willing to comment on this? -paul +------------------------------------------------------------------------+ | Prof. Paul A. Fishwick.... INTERNET: fishwick@bikini.cis.ufl.edu | | Dept. of Computer Science. UUCP: gatech!uflorida!fishwick | | Univ. of Florida.......... PHONE: (904)-335-8036 | | Bldg. CSE, Room 301....... FAX is available | | Gainesville, FL 32611..... | +------------------------------------------------------------------------+