mahout-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From unmesha sreeveni <unmeshab...@gmail.com>
Subject Re: Naive Bayes Classifier for Marketing data
Date Tue, 04 Feb 2014 13:51:22 GMT
Sorry.
But in defenitive guide i saw binning tags. I dnt know about whether
numerical binning is possible in mapreduce.


On Tue, Feb 4, 2014 at 7:09 PM, Ted Dunning <ted.dunning@gmail.com> wrote:

> Not to be rude, but how would you think you could use binning in a
> map-reduce program?
>
> Seriously, what ideas can you come up with for this?  What are the problems
> you see?
>
>
> On Tue, Feb 4, 2014 at 4:29 AM, unmesha sreeveni <unmeshabiju@gmail.com
> >wrote:
>
> > How to use binning in mapreduce
> >
> >
> > On Tue, Feb 4, 2014 at 2:59 PM, Ted Dunning <ted.dunning@gmail.com>
> wrote:
> >
> > > Mandeep,
> > >
> > > I just worked through a similar example using the same data set but
> using
> > > the logistic regression learner.
> > >
> > > In order to use Naive bayes, you would need to convert the continuous
> > > variables to categorical variables by binning.
> > >
> > >
> > >
> > > On Mon, Feb 3, 2014 at 11:03 PM, mandeep singh
> > > <mandeep.ma.singh@oracle.com>wrote:
> > >
> > > > Hi,
> > > >
> > > > Mahout In Action book says that naive Bayes strongly prefers
> text-like
> > > > data.
> > > > Is it possible or worth to run naive bayes classifier on the
> marketing
> > > > data in the following format, also mentioned in the book:
> > > > Data has these fields: age gender typeId colorId price discount
> > offerTime
> > > > purchaseDelay purchased
> > > > Similar data formats on which I want to run naive bayes:
> > > > http://archive.ics.uci.edu/ml/datasets/Bank+Marketing
> > > >
> > > > Thanks.
> > > >
> > >
> >
> >
> >
> > --
> > *Thanks & Regards*
> >
> > Unmesha Sreeveni U.B
> > Junior Developer
> >
> > http://www.unmeshasreeveni.blogspot.in/
> >
>



-- 
*Thanks & Regards*

Unmesha Sreeveni U.B
Junior Developer

http://www.unmeshasreeveni.blogspot.in/

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message