Re: assumption of Classification


Richard Ulrich wrote:
> On 1 May 2005 20:04:05 -0700, "Reef Fish"
> <Large_Nassau_Grouper@xxxxxxxxx> wrote:
>
> >
> > Richard Ulrich wrote:
> > > On 28 Apr 2005 21:05:34 -0700, "Reef Fish"
> > > <Large_Nassau_Grouper@xxxxxxxxx> wrote:
> > >
> > > >
> > > > Richard Ulrich wrote:
> > > > > On 26 Apr 2005 22:21:23 -0700, "Data Matter"
<fungile@xxxxxxxxx>
> > > > > wrote:,
> > > "Data Matter"
> > > - don't be concerned with Bob's rant about me.
> > >
> > > He did not contradict anything I said, and he added some detail.
> >
> > Really?
> [... ]
> Okay, more precisely: He did not contradict any ADVICE.
>
> Read my advice in my post, and not Bob's version of it.
>
> [... ]
>
> > Exhibit 2.

What happened to Exhibit 1? Exhibit 3? Exhibit 4?

Your post is just a self-serving obfuscation, even about Exhibit 2:

> > RU> normality is not the only assumption to be checked.
> > RU> It's always good to check.
> >
> > RF> WHY? If the clustering model assumes nothing about normality?
>
> Here is the original. I've corrected the attributions, and
> restored an important sentence.

Except you've DELETED what Data Matter said!!! I'll put it back for
you.
He said NORMALITY is NOT one of the assumptions.
>
> Data Matter >

DM > He's asking if these procedures make distributional assumptions.
DM > Classification trees do not. Most clustering algorithms (k-means,

DM > single link, average link, etc.) do not.

Read the TWO "do not" in the sentences above. Then Richard Ulrich
followed immediately with his bum advice I cited, that "It's always
good to check <normality>. That's why I asked, "WHY? If the
clustering model assumes nothing about normality?"

Let me repeat. Why is it a good idea to check for normality if
it's NOT assumed in the model?

Richard, you're not only ignorant in the subject but you are down
right DISHONEST in your attribution and self-serving "clever
editing" to complete distort the pertiment part of the discussion.


> > Nonetheless, normality is not the only assumption to be checked.
Every
> > method has its own list of assumptions and you should make sure
that
> > your data agree with the method you choose.
>
> RU > It's always good to check.
>
>
> [Snip exhibit 3 and 4... ]

Why were Exhibits 3 and 4 snipped?

Exhibit 3.

The Ulrich went into his characteristic tangent that had NOTHING to do
with clustering by referring to checking OLS Regression assumptions:

RU> For methods of ordinary least squares, normality is not
RU> as important as having decently behaved residuals

This was TANGENT to clustering, yet an error which I corrected,
hence contradicted what Ulrich falsely claimed above.


RF> The iid N(0, sigma^2) assumption about the ERRORS in a
RF> typical model
RF> < ...>. This is UNDERGRADUATE stuff, Richard, and you said
RF> "normality" is not as important ...".

Normality is NOT vs Normality IS. That's a contradiction.


Exhibit 4.

This is still a TANGENT issue unrelated to clustering, brought forth
by Richard Ulrich. Again, this time he contradicted HIMSELF about
the unimportance of normality in OLS Regression problems.

RU> absence of outliers,

RF> Absence of outliers with respect to WHAT? (Normality of course).

RF> And they DO NOT pertain to "Classification" in the sense of
RF> "Numerical Taxonomy", "Clustering", and various agglomerative
RF> and divisive methods (algorithms) commonly used in CLUSTERING.

Check absence of outlier vs do NOT pertain to clustering
is a contradiction.


So there! I exumed the THREE Exhibits Ultich swept under the rug,
showing how I corrected and CONTRADICTED his statements which he
self-servingly claimed that I merely "added comments" to his. LOL


> > I cited you verbatim and myself verbatim above, in the Exhibits.
> > Where had I not read you correctly?
>
> See above, for the easy one.
> - if anyone else thinks I "snipped" anything valid and useful,
> Exhibits 3 and 4, let me know.

I not only think, but I KNOW that you snipped everything valid and
useful (in terms of your ERROR and my contradiction of your error)
in Exhibits 2, 3, and 4.

I let the READERS know (see above) your mis-attribution,
misrepresentation, and self-serving dishonesty in addressing the
ISSUES which I discussed.

You had your say. I gave you my rebuttal. It's up to the READERS,
to see for themselves who's correct/honest/dishonest in their
summary of out exchanges.

I am not interested in any further bandwidth-wasting rebuttal of
all your false allegations and assertions on this topic.

Our discussion will be retrievable in groups.google.com for YEARS
and YEARS to come. Both our knowledge AND credibility can stand
on what we have written.

I end with this same advice to all, especially to Richard Ulrich:

"Stick to what you KNOW. And leave the BS to the barn in the
farm. That would be the greatest contribution you can give to
any statistical/mathematical readership, such as this one. "

-- Bob.

.