[SciPy-User] Mailing list?

Simon Friedberger simon+python@a-oben....
Sun Nov 22 12:46:21 CST 2009


Ok, apparently this message got through, so that answers my question.
Here is my original messages.
Sorry for the confusion.



Good Night Everybody,

I just looked at the documentation for the K-Means vector quantization
functions and I am a bit confused. On the one hand it says that
normalization to unit variance would be beneficial on the other hand
there are a lot of "must"s in the descriptions.
I was wondering if it is possible to use the functions without
normalization or if there is a negative impact.

This would make sense because it seems reasonable that one would want to
build the codebook on some set and then quantize a different set. In
this case normalization would have to be the same or be omitted.

I am also interested in literature recommendations concerning why this
is a good idea in general.

Any help would be greatly appreciated.

Best
Simon



On 19:42 Sun 22.11.09, Simon Friedberger wrote:
> Hi everybody,
> 
> I sent a message about the K-Means algorithm a couple of days ago but it
> seems like it never made it on the list. Are new members moderated or
> something?
> 
> Best
> Simon
> _______________________________________________
> SciPy-User mailing list
> SciPy-User@scipy.org
> http://mail.scipy.org/mailman/listinfo/scipy-user


More information about the SciPy-User mailing list