Hi Jamal,
It is to be noted that there is NO straight implementations
for Hierarchical clustering in Mahout.
But still you can be able to achieve Hierarchical clustering
through the techniques mentioned in the link,
You can either follow Top-Down / Bottom-Up approaches to do
the Hierarchical clustering.
As per the description it says initially you create 'K' top
clusters and then using *ClusterOutputPostProcessorDriver *you can copy
each clustered points to a separate folder and then again run the K-Means
on individual cluster then you can get the sub-clusters for the big one.
"Mahout in Action" book is the best reference book for
getting familiarity with Mahout along with some nice examples.
Best,
Mahesh Balija,
CalsoftLabs.
On Tue, Jan 29, 2013 at 4:11 AM, jamal sasha <[email protected]> wrote:
> Sorry.. accidental sent out:
> But as I was saying..
> I was looking in the link :
> https://cwiki.apache.org/confluence/display/MAHOUT/Top+Down+Clustering
> but its not very clear how to perform heirarchical clustering?
> Also, in the end.. I would also want to get the ids where each of the
> cluster center lies.
> Is there a way to do this in mahour?
>
>
> On Mon, Jan 28, 2013 at 2:40 PM, jamal sasha <[email protected]>
> wrote:
>
> > Hi,
> > I have data in format:
> > id::f1,f2,f3....fn
> > Now, I want to perform heirarchical clustering in this data
> > I am looking over here:
> > https://cwiki.apache.org/confluence/display/MAHOUT/Top+Down+Clustering
> >
>
