Don't swarm for Anomaly models

rhyolight · September 19, 2016, 4:42pm

Those values are “width” and “radius”:

numenta/nupic-legacy/blob/master/src/nupic/encoders/date.py#L48-L50


      
          
          :param season: (int | tuple) Season of the year, where units = day.

You can see exactly what is happening with those values in the date encoder:

github.com

numenta/nupic-legacy/blob/master/src/nupic/encoders/date.py#L101-L113


      
          if season != 0:
            # Ignore leapyear differences -- assume 366 days in a year
            # Radius = 91.5 days = length of season
            # Value is number of days since beginning of year (0 - 355)
            if hasattr(season, "__getitem__"):
              w = season[0]
              radius = season[1]
            else:
              w = season
              radius = 91.5
          
            self.seasonEncoder = ScalarEncoder(w = w, minval=0, maxval=366,
                                               radius=radius, periodic=True,

As you can see above, the width and radius are used directly to create a periodic scalar encoder. See also:

Topic		Replies	Views
Swarming for anomaly detection with multi-column datasets NuPIC swarming , anomaly-detection	3	1594	January 5, 2017
Swarming on anomaly detection mission NuPIC	2	728	August 5, 2019
Why am I seeing lot of false positives? NuPIC	12	2484	June 22, 2016
Some questions on Anomaly Detection NuPIC anomaly-detection	1	732	May 5, 2018
Bad Anomaly detection for complex periods data NuPIC usage-help , anomaly-detection	4	2052	October 10, 2019

Don't swarm for Anomaly models

Related topics