Classification: NPL Management Ltd - Public

Thank you Matt. Those links are very useful! 

 

I am working with John Blackburn at the National Physical Laboratory on
some data we have from sensors on a bridge. We were wondering if we can
get NuPIC to learn with the data for the first 6 months and then get it
just to continue to just predict or show anomalies for the remainder of
the data? Essentially, is it possible to turn learning off at some point
during a run?

 

Regards,

Cavan

 

From: nupic [mailto:[email protected]] On Behalf Of
Matthew Taylor
Sent: 18 August 2014 17:00
To: NuPIC general mailing list.
Subject: Re: [nupic-discuss] Swarming [NC]

 

Hi Cavan,

 

Our documentation on model parameters is not great, so I created this
ticket to track progress on filling it in: 
https://github.com/numenta/nupic/issues/1214.

 

You can find the docs we do have here: 
https://github.com/numenta/nupic/wiki/Models#modelparams-schema

 

Also, have you seen these wiki pages about swarms?

- https://github.com/numenta/nupic/wiki/Running-Swarms

- https://github.com/numenta/nupic/wiki/Swarming-Algorithm

 

Regards,




---------

Matt Taylor

OS Community Flag-Bearer

Numenta

 

On Fri, Aug 8, 2014 at 7:40 AM, Cavan Day-Lewis <
[email protected]> wrote:

Classification: NPL Management Ltd - Commercial

Hi,

 

I was wondering if someone could answer some questions I have about
swarming?

 

In the "rec_centre_hourly_model_params.py" file (using the hotgym
tutorial as an example) outputted after a swarm what do the parameters
it produces mean? What are they used for? And why are they important?
Also, without looking into the algorithms, what sort of calculations is
the swarm making? I understand It goes through the input data and goes
through different models and an optimization process (PSO) where it
analyses which model best suits the dataset. I also understand that the
swarm finishes when the error score no longer decreases, but how do you
generate the error score? What are the technical differences between the
different swarm sizes? Again, I know a bit about it but an explanation
would be great.

 

In particular, I would be very grateful if someone could explain what
the following parameters within the model_params.py file represent:

'clParams': { 'alpha':

'_classifierInput': {'n':

'spParams':{ 'synPermInactiveDec':

'tpParams': { 'activationThreshold':

'tpParams': { 'minThreshold':

'tpParams': { 'pamLength':

 

I have just been watching the following video and found it very helpful
but I still have some questions.

https://www.youtube.com/watch?v=xYPKjKQ4YZ0

 

Apologies if these questions have already been asked before, link me to
a feed if so.

 

Thanks in advance,

 

Cavan

 

Mr Cavan Day-Lewis

Vacation Student

National Physical Laboratory

Hampton Rd | Teddington | Middlesex | UK | TW11 0LW

t: 020 8943 XXXX

e: [email protected]

w: www.npl.co.uk

 

 

 

 

________________________________

If you have received this message in error, please notify us and remove
it from your system.
NPL Management Ltd cannot guarantee that the e-mail or any attachments
are free from viruses.

NPL Management Ltd is a company registered in England and Wales, number:
2937881
Registered office: Serco House | 16 Bartley Wood Business Park | Hook,
Hampshire | UK | RG27 9UY 

________________________________

  <http://www.npl.co.uk/> 

STAY CONNECTED
  <http://www.twitter.com/npl>   <http://www.facebook.com/npldigital>  
<http://www.youtube.com/npldigital>  
<http://itunes.apple.com/us/podcast/national-physical-laboratory/id29507
6289>   <https://ktn.innovateuk.org/web/measurement-network> 


_______________________________________________
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