I'm trying to provide prediction of one data based on other data. I.e. We observer how A and B happen in time. Then we want to predict B based on current A alone. I'm trying to figure if HTM is the right approach for associative training.
On Dec 18, 2013, at 11:42 AM, Matthew Taylor <[email protected]> wrote: > I don't think feeding them interchangeably is the right approach. That > would spread the temporal patterns out and split them up over time, > and I'm not sure the CLA will do well. Are you trying to provide > predictions for both inputs? > --------- > Matt Taylor > OS Community Flag-Bearer > Numenta > > > On Wed, Dec 18, 2013 at 9:21 AM, Dennis Stark <[email protected]> wrote: >> I was wondering how CLA can be used with multiple sensory inputs? >> >> I.e. I have visual input and auditory input (both converted to SDRs). Do I >> feed them interchangeably if I want to for association between two different >> stimuli and use one to predict the other. OR is it more along the idea of >> HTM? Where two CLAs would merge into a third CLA? Perhaps any other usage >> scenarios exist? >> >> Dennis Stark >> _______________________________________________ >> 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 _______________________________________________ nupic mailing list [email protected] http://lists.numenta.org/mailman/listinfo/nupic_lists.numenta.org
