Wikipedia:Articles for deletion/PCA applied to yield curves


 * The following discussion is an archived debate of the proposed deletion of the article below. Please do not modify it. Subsequent comments should be made on the appropriate discussion page (such as the article's talk page or in a deletion review).  No further edits should be made to this page.  

The result of the debate was delete Proto ///  type  12:04, 30 June 2006 (UTC)

PCA applied to yield curves
Prodded by User:Avraham as having no content or context, deprodded by User:Blotwell, adding a reference to what it's supposed to be about, but still no content (disregarding the figures, which should be in Yield curve, if anywhere) or context. Not expandable. &mdash; Arthur Rubin | (talk) 16:48, 24 June 2006 (UTC)
 * Delete per nom. Tevildo 16:55, 24 June 2006 (UTC)
 * Comment. PCA  (now that context has been provided) could be applied to a mathematical model of yield curves.  But no such model has been provided, and nothing more could be said other than that sentence.  &mdash; Arthur Rubin |  (talk) 17:09, 24 June 2006 (UTC)
 * Keep and mark as stub. Maybe someone will expand it one day. --Gabi S. 17:29, 24 June 2006 (UTC)
 * Delete. It's just three graphs. Limited context, this is a dicdef at best. Ifnord 19:11, 24 June 2006 (UTC)
 * Comment. The article does not make any sense to me. What is this "motion" thing? These curves are not wiggly worms, are they, that change shape as you look at them? What is on the horizontal axis of these curves? Isn't that time? So are there two kinds of time, a meta time for seeing how evolution in normal time evolves? And on whose authority does Wikipedia claim that shift, rotation and curvature are the three main movements? Rotation looks like pseudo scientific junk; it does not commute with linear scale transformations and can turn graphs of functions into non-functional graphs. "Curvature" does not appear to mean what curvature is supposed to mean. --Lambiam Talk 19:18, 24 June 2006 (UTC)
 * "What is on the horizontal axis of these curves?"--- exactly! What do the axes mean?  That is basically the problem with PCA; statisticians can't tell you what it means to rotate "options price" into "oil production" or whatever.  The geometry makes sense because mathematicians can deal with abstract euclidean space, and the euclidean metric happens to be consonant with elementary parametric statistics (e.g. Gaussians), but how is one to interpret a rotation (as in the principle axis theorem) outside the realm of euclidean geometry (i.e. independent of the original meaning of the axes)? ---CH 10:10, 26 June 2006 (UTC)


 * Delete, limited context, almost nonsensical. --Coredesat 21:06, 24 June 2006 (UTC)
 * Delete, per above. Oleg Alexandrov (talk) 02:12, 25 June 2006 (UTC)
 * Delete, but for different reasons - as presented it is original research, there are similar things that are done with yield curves, but right now I'm not up on them enough to properly edit this. The principal components stuff seems like a logical way of approaching this problem, but where is it from? Smallbones 14:28, 25 June 2006 (UTC)
 * Delete as nonsense vio of WP:NOR. Actually, this vapid stub is a fitting :-/ parody of PCA, a widely used but suspect method in statistics (the geometry is fine; the problem is with the interpretation; as typically used, PCA is just a method for lying with statistics--- in a way, it's rather fascinating that such a thing is possible, but plainly I digress.)  ---CH 10:06, 26 June 2006 (UTC)
 * People, people: you and I may be trained as mathematicians: we know what "rotation" and "curvature" are "supposed to mean", and we beat anyone with a big stick who is so sloppy as to use it to mean anything else, but it seems clear enough to me from the context that by rotation is meant "vertical shear" (which certainly does take functions to functions) and by "curvature" is meant some quadratic term. Here is just one of the many Google hits for PCA "yield curve": note that it implies that yield curves are the important application of PCA in finance.  Keep.  —Blotwell 10:36, 27 June 2006 (UTC)
 * No, I said rotation and I meant rotation; I just didn't take the time to try explain in detail. Keep googling, if you are curious, and you'll find my critique of PCA from several years ago. ---CH 02:45, 28 June 2006 (UTC)


 * Comment. Even Blotwell's comment is correct, there's nothing correct in the article at present.  Delete, and start over.  &mdash; Arthur Rubin |  (talk) 20:56, 27 June 2006 (UTC)
 * The above discussion is preserved as an archive of the debate. Please do not modify it. Subsequent comments should be made on the appropriate discussion page (such as the article's talk page or in a deletion review). No further edits should be made to this page.