machine brains

Louis Savain savainl at earthlink.net
Thu Nov 13 12:40:11 EST 1997


In article <64e14a$elu at ux.cs.niu.edu>, rickert at cs.niu.edu (Neil
Rickert) wrote:

>In <01bcefc9$33b20160$e67a61ce at asdf> "ray scanlon" <rscanlon at wsg.net> writes:
>>Neil Rickert <rickert at cs.niu.edu> wrote in article
>><64a8qu$abu at ux.cs.niu.edu>...
>>> In <01bceec8$7cb77900$fd7a61ce at asdf> "ray scanlon" <rscanlon at wsg.net>
>>writes:
>
>[...]
>
>>What is needed is the ability to generalize, the ability to clear off
>>some trees so we may have a clear view of the forest.
>
>But I think you have made it very difficult to generalize by
>insisting on a bottom up neural net methodology.

  I may be misinterpreting your message but isn't it the top down
approach that makes it difficult to generalize?  Doesn't approaching
the problem using simple components much more in keeping with general
emergent behavior?

Regards,

Louis Savain

President, Marengo Media, Inc.



More information about the Neur-sci mailing list