Baptize my method
June 22, 2005 9:02 AM   Subscribe

NameFilter: Please help me give a name to a method I developed ....

It is based on decorrelation scales and we use it to fill in gaps (missing values) in datasets. The paper is coming out soon but I am missing a name that will stick. Nothing fancy, please... I cannot come up with anything better than Decorrelation Based Filling Method (DBFM)?
I promise to aknowledge you properly...
Thanks!
posted by carmina to Science & Nature (26 answers total)
 
Decorrelative Completion Method (DCM)? Anything with ??FM is bound to be misread * * F'ing Manual. DCM has the added benefit of being my initials :)
posted by kc0dxh at 9:29 AM on June 22, 2005


Why not just name it after yourself? The [Your Name Here] Method? I can't think of anything memorable that would stick when it comes to decorrelation scales and missing values in datasets. "The Carmina Method".

I want a quarter everytime somebody says it ;-)
posted by Servo5678 at 9:30 AM on June 22, 2005


It might help to know if your method applies to cryptography, audio processing, mapping/graphical processing or another field -- that way, we can try to come up with something catchy and apropros.
posted by Merdryn at 9:45 AM on June 22, 2005


Data Value Interpolation Using Decorrelation ?

Data Set Interpolation By Decorrelation ?
posted by orthogonality at 9:54 AM on June 22, 2005


Gapbuster.
FillIt.
The UnHoley Data Munger (of Doom).
Sparse-B-Gone.
posted by Caviar at 9:59 AM on June 22, 2005


You can get rid of the 'Based' in DBFM; just DFM would suffice.
Decorrelation Interpolation (DI) has a nice ring to it. Or if you can spell anything with the initials of the people working on the paper (if there's more than just you).
posted by fleacircus at 10:25 AM on June 22, 2005


Destroyer of Worlds.

No, I'm actually serious. That's not a snark. I genuinely think you should name it that, because people would take notice.

However, if you choose to ignore my sage advice, then perhaps the DCM or DFM suggestions above would suit your needs. I'd say stick with three-letter acronyms, if you choose to go the acronym route.
posted by aramaic at 10:37 AM on June 22, 2005


If you name it DFM, you'll have lots of grizzled old military types thinking that your technique "doesn't fucking matter." That may be important to you.

I too would like to know the field before I take a shot, but so far I like fleacircus's idea the best.
posted by patgas at 11:14 AM on June 22, 2005


I wouldn't use the method 'M' in your acronym - look at what happened to "ATM Machine" and "PIN Number". Given a little bit of time, people will be calling it the "DFM Method".

I second the idea of using your name/your team name/your institution name as the method name.

"The Minnesota Method", for instance. Nice ring to it. In the context, it will always be understood.
posted by unixrat at 11:14 AM on June 22, 2005


Usually that wouldn't be "filling" but "imputation," I think.

Decorrelation Imputation Method --> DIM. oops.
Decorrelation Imputation Scheme --> DIS.
Imputation By Decorrelation --> IBD.
Imputation Method Using Decorrelation --> I-MUD.
Decorrelation-Based Imputation Method --> D-BIM.
Decorrelation Based Imputation Scheme --> D-BIS

Is this similar to gary king's AMELIA?
posted by ROU_Xenophobe at 11:25 AM on June 22, 2005


(I might mean AMILIA)
posted by ROU_Xenophobe at 11:25 AM on June 22, 2005


How about the FIG (Fill In Gaps) method? If I got to name something, and didn't want to name it after myself or my girlfriend, it'd name it something funny and memorable.
posted by pwb503 at 11:54 AM on June 22, 2005


yes, I don't feel very comfortable with my name on it... Keep it coming people! great ideas... thanks!

Rou- the idea is similar: filling in missing data from large datasets. However, AMELIA (not AMILIA), as I just googled it,
seems to be different than our method. We use higher order statistics. Their method would be more appropriate, I guess,
for smaller datasets (such as those used in political science or medical research). I am looking at very large datasets coming in from satellites and such. Thank you for the reference though.

patgas, you have just sort of ruined DFM now for me... :-)
posted by carmina at 12:32 PM on June 22, 2005


Find Unknown Correlation Komputation?

Correlation Replacement Authentication Procedure?

Decorrelation Orientation Reversal Korrelation?
posted by cincidog at 12:44 PM on June 22, 2005


How about the Carmina method? Methods/algorithms/formulae are named after the inventor all the time.
posted by borkencode at 12:52 PM on June 22, 2005


Why not just Decorrelative Filling (DF) or Decorrelative Interpolation (DI)? Short and very versatile since it's easy to conjugate from noun to verb. Adding "Method" is unnecessary language and makes it clunky to use in conversation.

For posterity, add a pronoun (your name, state, child's name, favorite author/musician/architect/artist) in front:
cDF (Carmina Decorrelative Filling)
cDI (Carmina Decorrelative Interpolation)

posted by junesix at 1:52 PM on June 22, 2005


Methods/algorithms/formulae are named after the inventor all the time.

But usually someone else names them after the inventor, and that's often after the inventor is dead or retired. It definitely wouldn't do in academia to actively name something after yourself. Besides, what if there are problems with the method etc.? Then you have people writing papers attacking the Carmina method, and that's no fun.
posted by advil at 2:29 PM on June 22, 2005


advil, I hear you. If the method catches up (really it is no big shit, just another --of the hundred-- ways to do it). I seriously do not want my name on for several reasons. But I would like an acronym for quick reference inside my own paper and in other areas where it might come up... It is actually customary to baptize your own method in this field....
posted by carmina at 2:41 PM on June 22, 2005


Merdryn, it is a method to fill in missing values from satellite coverage of the earth and measurements of fluxes at the surface.

Caviar and aramaic, yeah, I wish we were that coooll in this field. I somehow feel the editors will piss on me little paper...
Although, they will congratulate me in person afterwards. Go figure!

unixrat, I will keep that in mind, thnx.

and to the rest, foul-mouthed darlings.... what can I say... well, I guess I asked for it... smooch!
posted by carmina at 2:53 PM on June 22, 2005


One word: strategalutions
posted by Who_Am_I at 3:16 PM on June 22, 2005


Quite obviously you have invented the Flux Capacitor.
posted by ajp at 3:54 PM on June 22, 2005


Decorrelation Interpolation for Large Datasets .. hrm, no.
Filling Using Decorrelation?

more seriously, maybe
Interpolation by Decorrelation (IBD)?
posted by fleacircus at 4:59 PM on June 22, 2005


If the method catches up (really it is no big shit, just another --of the hundred-- ways to do it). I seriously do not want my name on for several reasons.

Take part in a time honored unix tradition: Yet Another Decorrelation Completion Method.
posted by boo_radley at 5:31 PM on June 22, 2005


Higher order statistical methods for decorrelated interpolation of data? (HOSMID)? (Or just SMID?)
posted by WestCoaster at 8:38 PM on June 22, 2005


How about Dataset Entry Fill Tasker (DEFT)? Unwieldy at first, but a great association thereafter.

If not, perhaps Dataset Fill Provider (DFP)? Whatever you decide, run it through acronym finder. There are a few amusing entries for DFP and I'm sure others suggested, but it's tough to find a three-letter acronym that doesn't have at least one. DEFT, on the other hand, is pretty clean (well, except for Do Every F*cking Thing, but I'd like that association if I were you!).
posted by melissa may at 9:34 PM on June 22, 2005


melissa, thanks for the site!

thank you kids. it was great help.

I decided to go with DBI (decorrelation based interpolation) which melissa's acronym finder was pretty kind to. Let the damn thing go now!
posted by carmina at 5:37 AM on June 23, 2005


« Older Soundcard issue   |   Graphical bookshelf for etexts? Newer »
This thread is closed to new comments.