Said Tel Aviv's Yo'v Binyamini,
For tests si-mul-ta-ne-ous-li-ni.
So don't go too far,
Just use FDR.
Those errors will stay teeny weeny.Raft'ry used model to cluster,
Weird decomposition he'd muster.
Sorts pods versus peas.
Naes replied, "Oh, Please
Criterion-Based, you can trust her."
A passionate Friedman quite smart.
Pursues better methods: MARS, MART.
With zillions of trees,
It sounds like a breeze.
But what's next? A system named TART?
To Gordon at Williams we ran.
DeVeaux said that it would be grand.
The heat makes us swell.
The gnats were from hell.
And wine with our dinner was banned!
Chorus:
That was a very fine verse.
Sing me another one.
We're having lots of fun.
GRC's No.1.
GRC, GRC, GRC.
Kim's lecture was image-i-nary,
And showed us some pictures quite scary.
His banana went brown,
Which caused him to frown.
His yogurt then looked slightly hairy.
Our lim'rick man is Dennis Keller.
He brings us humor from the cellar.
His talent is great,
And doesn't run late.
Which makes him a popular feller.
There was a young man named Bert Gunter.
At Merck a statistical punter.
When asked for his name,
He said "'twas the same,
To say something else is a blunder."
Re-con-ci-li-a-tion of data,
A node-worthy hot baked potato.
But error's is gross,
And not really toast,
So we'll have to update you later.
An Argentine techie, Miguel.
He reconciled data so well.
My eye was distracted,
The lecture detracted
By his laser pointer from hell.
W P G n b
I m S C h
r e F
It's all Greek to Me!
t D Y a z q
[Chorus] |
|
A Bayes-Statistician from Kent
Had trouble with his "component."
He was in a hurry,
But said, "Don't you worry,
I know it's somewhere, though latent."A man can have great health because
He eats pizza with tomato sauce
But Karen says, "NO"
She undoubtedly knows
To correlate is not to cause.
Jerry told us how we could kill SPAM.
Junk mails for which we don't give a damn!
We lose some emails
When his filter fails
Your note got trashed too---what's a sham!
[Chorus]
There once was a stat man named Bayes,
Who mooned in peculiar ways.
He'd show his prior
To any admirer,
But to see his posterior, you pays!
We heard about Bayesian interiors.
That some people think are superiors.
This called into question,
The age old suggestion,
That Bayesian have better posteriors.
Our friend Phil Brown wants us to shrink
To make good decision we think.
Diffuse priors are key,
You surely must see,
Or we'll drive Marina to drink.
[Chorus]
Our Speaker on Thursday is Gerry.
His hairline is getting more "airy".
He brought some Rogaine
To thicken his mane.
But Keller's a user---Be wary!
Our chairman gets data from spies.
He'll make stuff up and tell us lies
On Thursday retorts,
"So just eat my shorts,
With mayo and mustard and flies."
Some people devote special care.
To boosting and pruning their hair.
For others, like Keller,
And that Ungar feller,
That field of endeavor is bare.
FDR, MBC, CBC,
PCA, PLS, AMT,
MARS, MART, MIA, MIR,
Bags, Boosting, Oh Dear!
CART, MAR, MIX, MID, GLIM, GLIMPSE, SPLAT, U.D.
[Chorus] |