Indeed we have a long way to go, it did however lead my thoughts in the 
direction of some kind of noise filter algorithm that enables a more efficient 
memory usage in long term learning prediction systems.
I know that most people today do training and then disable it to allow for 
anomaly checking and/or prediction.

Well, I will give it a go and see if anything interesting will arise out of my 
efforts. If that happens we can pick up the discussion again.


-Hannu







On 01 Nov 2013, at 19:04, Jeff Hawkins <[email protected]> wrote:

> No this hasn't been discussed here.  As you say, it is a meta conversation.
> If there are others who want to discuss this kind of research you might want
> to start a separate email thread as it is pretty far removed from most of
> what NuPIC is about.  There is a big gap between understanding how ensembles
> of neurons process information and measured IQ in living humans.  Obviously
> psychology (and IQ) are based on neural structures but to me we have a long
> way to go before we can unite these two fields.
> 
> Occasionally you can find some connections.  For example some scientists
> argue that dyslexia is simply a problem with slow processing in neural
> tissue.  The left side of the cortex, where language is, is more heavily
> myelinated than the right side of the cortex.  Mylenation makes action
> potentials travel faster.  If you slow down speech, a dyslexic child starts
> to understand more normally.  The speculation is that poor myelination in
> the left hemisphere could cause dyslexia.  This is an idea that bridges low
> level neuroscience with high level intelligent phenomena.  These kind of
> examples are rare though.
> 
> Jeff
> 
> -----Original Message-----
> From: nupic [mailto:[email protected]] On Behalf Of Hannu
> Kettinen
> Sent: Friday, November 01, 2013 6:30 AM
> To: NuPIC general mailing list.
> Subject: [nupic-dev] A Strong Interactive Link between Sensory
> Discriminations and Intelligence
> 
> Folks, Has this been discussed here before? And if not, what are your
> thoughts on this?
> 
> source: http://www.bcs.rochester.edu/people/Duje/papers/13_Melnick_IQ_CB.pdf
> 
> It all started with a question, why do some people have higher IQ than
> others, I have my own theories but to me this paper is very exciting.
> If we can identify the factors that separate us IQ wise, we should be able
> to tweak those factors to boost IQ. (virtual or real for what it's worth).
> 
> I know this is a meta discussion for this ml, but I am just wondering if
> anyone else have been looking at the issues/possibilities from this kind of
> perspective?
> 
> Personally I am not worried about the technicals behind it, because the path
> Jeff is on can be combined with certain other features and algorithms to
> create a solid scalable prediction framework with capabilities far from
> 'just' prediction and anomaly detection in temporal sensory input.
> 
> 
> 
> 
> 
> 
> 
> 
> _______________________________________________
> nupic mailing list
> [email protected]
> http://lists.numenta.org/mailman/listinfo/nupic_lists.numenta.org
> 
> 
> _______________________________________________
> nupic mailing list
> [email protected]
> http://lists.numenta.org/mailman/listinfo/nupic_lists.numenta.org


_______________________________________________
nupic mailing list
[email protected]
http://lists.numenta.org/mailman/listinfo/nupic_lists.numenta.org

Reply via email to