Hi David, Thanks for the quick answer! Is it then correct to think of the NuPIC GespatialEncoder as a scheme of segmenting the 3D space into a set of small boxes (as a generalization of squares in the 2 dimensions)?
Regards, Dionysis On 27 November 2015 at 13:50, David Ray <[email protected]> wrote: > Hi Dionysis, > > Welcome! Yes. The NuPIC GeospatialEncoder can use either 2 or 3 dimensions. > > Sent from my iPhone > > > On Nov 27, 2015, at 5:22 AM, Dionysis Manousakas <[email protected]> > wrote: > > > > Dear all, > > > > I am wondering whether the current implementation of nupic geospatial > > takes into account the altitude. I can see that the altitude is parsed > when > > the raw data are read. However, the description of geospatial anomaly > > detector and the tutorial on using nupic geospatial both describe the > > process only for 2D data. Thus, my question is whether the anomaly > > detector is directly applied on the 3 dimensions, or rather the altitude > > is omitted. > > > > Kind regards, > > Dionysis > >
