On Jun 17, 2009, at 2:51 AM, Ted Dunning wrote:
A principled approach to cluster evaluation is to measure how well the
cluster membership captures the structure of unseen data. A natural
measure
for this is to measure how much of the entropy of the data is
captured by
cluster membership. For k-means and its natural L_2 metric, the
natural
cluster quality metric is the squared distance from the nearest
centroid
adjusted by the log_2 of the number of clusters. This can be
compared to
the squared magnitude of the original data or the squared deviation
from the
centroid for all of the data. The idea is that you are changing the
representation of the data by allocating some of the bits in your
original
representation to represent which cluster each point is in. If
those bits
aren't made up by the residue being small then your clustering is
making a
bad trade-off.
In the past, I have used other more heuristic measures as well. One
of the
key characteristics that I would like to see out of a clustering is
a degree
of stability. Thus, I look at the fractions of points that are
assigned to
each cluster or the distribution of distances from the cluster
centroid.
These values should be relatively stable when applied to held-out
data.
For text, you can actually compute perplexity which measures how well
cluster membership predicts what words are used. This is nice
because you
don't have to worry about the entropy of real valued numbers.
OK, so how do we go about codifying this stuff? Is there existing
code that we could use or is it worth us writing our own?
Some references would be good here, too. Feel free to add to http://cwiki.apache.org/confluence/display/MAHOUT/ClusteringYourData
. (I've already linked this conversation, but will probably cut and
paste some of it too.
Manual inspection and the so-called laugh test is also important.
The idea
is that the results should not be so ludicrous as to make you laugh.
Unfortunately, it is pretty easy to kid yourself into thinking your
system
is working using this kind of inspection. The problem is that we
are too
good at seeing (making up) patterns.
I think this is where the new Open Relevance Project can come in,
too. Judgments, etc. ain't just for search!
On Tue, Jun 16, 2009 at 2:35 PM, Grant Ingersoll
<[email protected]>wrote:
What tools/approaches are people using to validate their clustering
output?
Are there utilities that we should be implementing that would make
this
easier for users?
--------------------------
Grant Ingersoll
http://www.lucidimagination.com/
Search the Lucene ecosystem (Lucene/Solr/Nutch/Mahout/Tika/Droids)
using Solr/Lucene:
http://www.lucidimagination.com/search
