A Temporal Connectionist Model of Perceptual and Motor Learning

Rick Craik NOSPAMrcraik at ntl.sympatico.ca
Mon Jan 15 10:59:39 EST 2001


Mentifex wrote in message <93uv3c$br7$1 at nnrp1.deja.com>...
>In a searing Usenet performance, Louis Savain wrote:
>> http://home1.gte.net/res02khr/index.htm
>>
>> This is the first installment in a series of articles on a new
>> model of intelligence based entirely on the timing of events.
>> More to come.
>>
>> MOBE
>Mentifex:
>... whatever "MOBE" stands for; but the indicated URL resolves
>to our long-standing fellow AI Usenetter Louis Savain -- bon jour!
>
>Nunc ad rem:  Even if the logical gating in the mind-model is
>based _somewhat_ on the timing of events, there still needs to
>be a schema of the routing or data-flow within the proposed mind.


Timing would be the cue on where to bundle signals as data.
A set of live signals start off at the same time is still not enough
to create a bundle. The attempt at bundling would have to
have placeholders for signals that are not firing at that particular
instance. With such a bundle of placeholders we can invoke the
term data, otherwise they are just timing signals. Without the
concept of data channel, we don't have the inception of formal
Information.

The bundling of signals is easy to achieve at external sensors.
But to attempt a schema farther in, bundling get's harder;
we may find feedback,  feedforward, or control signals. These
signals may be internal to a "data processing module" or
may be trying to flow to another module. So the question
may be, where are the "interconnects"?

Layering of nets is also a clue to bundling.

Regards,
Rick









More information about the Neur-sci mailing list