Thanks, for response, I'll will try this algorithm. -- Kumari Shalini
On Mon, Jan 18, 2016 at 10:40 AM, Divyang Shah via nupic < [email protected]> wrote: > > > ---------- Forwarded message ---------- > From: Divyang Shah <[email protected]> > To: "NuPIC general mailing list." <[email protected]> > Cc: > Date: Mon, 18 Jan 2016 05:09:17 +0000 (UTC) > Subject: Re: How to code > Hi kumari, > I think for categorization purpose better option is naive bayes algorithm. > > Regards, > Divyang Shah > > > On Saturday, 16 January 2016 9:38 PM, Kumari Shalini < > [email protected]> wrote: > > > My goal is just to classify data according to category. I have category , > and then I have to classify data which I have in my database to which > category data belongs. > > -- > Kumari Shalini > > On Sat, Jan 16, 2016 at 3:12 AM, Matthew Taylor <[email protected]> wrote: > > Kumari, > > You have to think about how you can transform your data into temporal > data. HTML data is spatial in nature. Are you going to be polling for HTML > documents over time? Are there any stats you can collect about the > documents over time that change significantly? > > A better question... what exactly is your goal? > > Regards, > > --------- > Matt Taylor > OS Community Flag-Bearer > Numenta > > On Fri, Jan 15, 2016 at 11:45 AM, Kumari Shalini <[email protected]> > wrote: > > Hi, > > I have some questions to ask? How will I get to know the data type like > scalar and how to know the dimensions of data? > > My data is a raw html data and my data is fully present at a point of time > , so how to use standard classification method? > > -- > Kumari Shalini > > On Tue, Jan 12, 2016 at 1:19 AM, Chandan Maruthi < > [email protected]> wrote: > > Kumari, > I think your problem/usecase here is that of classification. Lookup > classification methods in data science/AI to help your case. > You would use Nupic if your data has a temporal [time dependent] pattern > in it, For example notes of a song, temperature during a day etc. > > If you pattern is not temporal. and is represented fully by the data > present at a point of time then you can use standard classification methods. > > Regards > Chandan > > > Regards > Chandan > > On Mon, Jan 11, 2016 at 7:45 AM, Kumari Shalini <[email protected]> > wrote: > > Hi, > > I have gone through the links and video and understand what is numenta and > what is the use. So, I want to know to how to write code for encoding data. > For example- I have huge amount of data and a category to which the data > belongs. > Now, I have some some more data coming and now I have to predict to which > category the data belongs, Then how to do this? Please help me in > understanding . > > What Im understanding is I have to convert the category and data to SDR > and then convert the coming data to SDR then I have to match with all the > category SDR, to which itmatches more the data belongs to that category. Am > i right? > If I'm not correct then please correct me. > > Now, I want to know Is there any documentation to how to write code to > encode data. > I have to write code in python. > > Please help me. > > -- > Kumari Shalini > > > > > -- > Regards > Chandan Maruthi > > > > > > > >
