YKY --
About neural nets versus symbolic approaches: I am not the only one to
have long ago concluded that this is a bogus dichotomy
In both Novamente and Joshua Blue, knowledge is represented in a
semantic network in which nodes and links have
-- truth values, indicating degrees of truth
-- short and long term importance values, indicating the usefulness of
the node/link to the mind on different time scales
The importance values are updated by a specific equation and are
roughly analogous to time-averages of NN activation.
Using this kind of representation allows aspects of semantic nets and
attractor neural nets to exist in the same knowledge network.
There are many critical details that need to be gotten right to make
this kind of approach work, but the point is that NNs and logic-based
systems don't have to be dichotomized, because "semantic net" and
"attractor neural net" can be viewed as aspects of the same system...
-- Ben
On 6/13/06, Yan King Yin <[EMAIL PROTECTED]> wrote:
> What I said in my previous reply was that something very like neural nets
> (with all the beneficial features for which people got interested in NNs
in
> the first place) *can* do syntax, and all forms of abstract
representation.
>
> I do not think it is fair to say that they can't, only that the
> particularly restrictive interpretation of NN that prevails in the
> literature can't.
Hi Richard
I have to agree that NN can represent all forms of knowledge, since our
brains are NNs. But figuring out how to do that in artificial systems must
be pretty difficult. I should also mention Ron Sun's work, he has long
tried to reconcile neural and symbolic processing. I studied NNs/ANNs for
some time, but I recently switched camp to the more symbolic side.
One question is whether there is some definite advantage to using NNs
instead of say, predicate logic. Can you give an example of a thought, or a
line of inference, etc, that the NN-type representation is particularly
suited? And that has a advantage over the predicate logic representation?
John McCarthy proposed that predicate logic can represent 'almost'
everything.
If NN-type representation is not necessarily required, then we should
naturally use symbolic/logic representations since they are so much more
convenient to program and to run on von Neumann hardware.
YKY ________________________________
To unsubscribe, change your address, or temporarily deactivate your
subscription, please go to
http://v2.listbox.com/member/[EMAIL PROTECTED]
-------
To unsubscribe, change your address, or temporarily deactivate your subscription,
please go to http://v2.listbox.com/member/[EMAIL PROTECTED]
