Hello all, How is the algorithm able to choose the box size when reading the data in real-time, since the extrema are not known a priori?
Best regards, Dionysis On 11 December 2015 at 01:42, Matthew Taylor <[email protected]> wrote: > NuPIC can absolutely feed in real-time data. If you look a bit closer > at the examples that read data from a CSV file, many of them pass each > row into an OPF model object, which is what you could do yourself. > Just get your data whoever you like and pass in one row at a time into > the model. For example: > > https://github.com/numenta/nupic/blob/master/examples/opf/clients/hotgym/simple/hotgym.py#L77-L91 > --------- > Matt Taylor > OS Community Flag-Bearer > Numenta > > > On Mon, Dec 7, 2015 at 2:57 AM, shashank manjunath > <[email protected]> wrote: > > Hello! > > I'm a newbie here and have been going through the nupic documentation. > > I tried running the Hot-gym tutorial and was successful with it. > > My purpose of using Nupic is to develop an application which requires me > to > > take in real time data and analyse it using nupic algorithms and suitably > > perform an action( An alert or a notification). > > From what I have observed, the data being used currently in many examples > > are in the form of a csv file, basically a stored data-set. > > I wanted to know is there a way I can use Nupic to directly read real > time > > data and perform analysis on it. > > Also my other question is regarding the hot-gym tutorial where the > data-set > > used has inputs from every single day. Is it possible for us to use it in > > scenarios where we have a burst of data at only certain time period of a > day > > and it each day we may or we may not get the data. Can we apply the nupic > > algorithm in such a scenario? > > Please clarify the above. > > > > Regards, > > Shashank > > > >
