> From: Walker, Guy <ghw296@soton.ac.uk>
>
> Am I correct in thinking that symbolic computation has a
> distinct frame problem, and connectionist/neural networks
> do not? I am basing this on the assumption that things like
> recurrent, feed foreward/back neural networks can learn and
> adapt much more easily than is the case with symbolic
> computation.
You are right, but nets clearly can't do a lot of other thought-like
things that we can do (e.g. calculate, infer, speak, respond to speach,
solve problems, etc.) that symbol systems CAN do.
This archive was generated by hypermail 2b30 : Tue Feb 13 2001 - 16:23:51 GMT