mahout-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Daniel McEnnis <dmcen...@gmail.com>
Subject Re: Spectral Clustering, EigenCuts and Affinity Matrix
Date Mon, 11 Apr 2011 02:50:28 GMT
Lance,

My reading of it is that it is not a Wikipedia affinity matrix but a
distance matrix.  Representing a distance matrix as a graph,
clustering, then projecting back to a vector space is common for
bleeding edge (circa 2008) clustering algorithms.  It sounds like what
your describing in the email.

Daniel.

On Sun, Apr 10, 2011 at 7:09 PM, Lance Norskog <goksron@gmail.com> wrote:
> From the Mahout wiki:
> https://cwiki.apache.org/MAHOUT/spectral-clustering.html
>
> "This n by n comparison of all objects with all others forms the
> affinity matrix, which can be intuitively thought of as a rough
> representation of an underlying undirected, weighted, and
> fully-connected graph whose edges express the relative relationships,
> or affinities, between each pair of objects in the original data. "
>
> http://en.wikipedia.org/wiki/Affinity_%28mathematics%29
> Wikipedia defines 'affinity matrix' as an affine transform. Does the
> "affinity matrix" in Eigencuts match the Wikipedia entry, or is it a
> completely different concept that happens to be called an 'affinity
> matrix'?
>
>
>
> --
> Lance Norskog
> goksron@gmail.com
>

Mime
View raw message