ok, then the window would be the size of a set of values over which you compute the error. Ie you have s stream of data, and the error is computed over actual data (T) and all data to T-1000. The reason is smoothing, see "sliding average" for other example.
Cheers, Mark On Tue, Apr 15, 2014 at 10:58 AM, Sergey Bryukov <[email protected]> wrote: > Let's assume hotgym example.. > > _METRIC_SPECS = ( > MetricSpec(field='P0', metric='multiStep', > inferenceElement='multiStepBestPredictions', > params={'errorMetric': 'aae',* 'window': 1000*, 'steps': > 5}), > > > > > On 15.04.2014 12:40, Marek Otahal wrote: > > can you provide more context? What "program", file? > > > On Tue, Apr 15, 2014 at 10:38 AM, Sergey Bryukov <[email protected]>wrote: > >> What is param 'window' certain meaning? >> >> >> >> _______________________________________________ >> nupic mailing list >> [email protected] >> http://lists.numenta.org/mailman/listinfo/nupic_lists.numenta.org >> > > > > -- > Marek Otahal :o) > > > _______________________________________________ > nupic mailing > [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 > > -- Marek Otahal :o)
_______________________________________________ nupic mailing list [email protected] http://lists.numenta.org/mailman/listinfo/nupic_lists.numenta.org
