Re: Structured Programming using Forth
- From: Albert van der Horst <albert@xxxxxxxxxxxxxxxxxx>
- Date: 09 Apr 2007 22:02:52 GMT
In article <1176135579.663471.73680@xxxxxxxxxxxxxxxxxxxxxxxxxxxx>,
rickman <gnuarm@xxxxxxxxx> wrote:
On Apr 8, 3:34 am, "Jean-Fran=E7ois Michaud" <come...@xxxxxxxxxxx>
wrote:
The new SEAForth processor architecture is closer to a usable
artificial neural network than anything else ever created before, but
Regards
Jean-Francois Michaud
What aspect of the Seaforth chip makes it "closer to a usable
artificial neural network than anything else ever created before"? I
have seen many processors before with the same or even better
communications capability, for example, the transputer. I don't see
how this part is any better suited to neural networks than any other
device.
One of the aspects of natural neural nets is their resilience.
You can cut and reconnect, and they sort out how to get the
work done again.
Communication between transputers is brittle (I talk from experience)
and it remains to be seen whether Seaforth can do better.
NN are somewhere in between. If they are configured non-optimally
they can recover with some extra learning.
Groetjes Albert.
--
--
Albert van der Horst, UTRECHT,THE NETHERLANDS
Economic growth -- like all pyramid schemes -- ultimately falters.
albert@spe&ar&c.xs4all.nl &=n http://home.hccnet.nl/a.w.m.van.der.horst
.
- Follow-Ups:
- Re: Structured Programming using Forth
- From: Jean-François Michaud
- Re: Structured Programming using Forth
- References:
- Re: Structured Programming using Forth
- From: rickman
- Re: Structured Programming using Forth
- From: rickman
- Re: Structured Programming using Forth
- From: rickman
- Re: Structured Programming using Forth
- Prev by Date: Re: Forth Frustrations
- Next by Date: Minor Correction (was Re: Structured Programming using Forth)
- Previous by thread: Re: Structured Programming using Forth
- Next by thread: Re: Structured Programming using Forth
- Index(es):
Relevant Pages
|