Talk:Nonparametric regression

From Wikipedia, the free encyclopedia
Jump to: navigation, search
WikiProject Mathematics (Rated C-class, High-importance)
WikiProject Mathematics
This article is within the scope of WikiProject Mathematics, a collaborative effort to improve the coverage of Mathematics on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.
Mathematics rating:
C Class
High Importance
 Field:  Probability and statistics
WikiProject Statistics (Rated C-class, Top-importance)
WikiProject icon

This article is within the scope of the WikiProject Statistics, a collaborative effort to improve the coverage of statistics on Wikipedia. If you would like to participate, please visit the project page or join the discussion.

C-Class article C  This article has been rated as C-Class on the quality scale.
 Top  This article has been rated as Top-importance on the importance scale.

Proposed merge[edit]

I've proposed that "nonparametric multiplicative regression" be merged into this article. Here's why.

  • A Google search on "nonparametric regression" generates 369,000 hits, including dozens of book titles.
  • A Google search on "nonparametric multiplicative regression" generates 63 hits, most of which seem to be related to this paper, which describes a software package (apparently used by ecologists) in language like this: "a Gaussian (hump-shaped) curve ..." Which is fine, I guess, if you're an ecologist or a zoologist.
  • Clearly, NPMR is a subset of NPR, and should be subsumed under the larger topic. DavidCBryant 20:13, 27 February 2007 (UTC)

Dr. Hyndman's comment on this article[edit]

Dr. Hyndman has reviewed this Wikipedia page, and provided us with the following comments to improve its quality:

Methods described are incomplete, inadequately described, and do not cover some of the main methods (e.g., smoothing splines, regression splines, SVR, etc.)

NPMR is described in detail, even though this is not a particularly popular approach. Other widely used approaches are not mentioned at all.

The description of GPR/kriging is very brief, and refers to technical details that are not explained.

Kernel regression is very badly described. The data points are not convolved with a kernel function --- that is kernel density estimation. Rather, a weighted average of the dependent variable is taken where the weights are based on a kernel function applied to the independent varaibles.

A popular generalization of kernel regression is local linear (or local polynomial) regression including loess. This is not discussed at all.

GLMs are not a nonparametric regression method unless the regressors are splines, or some other general class of flexible basis functions.

We hope Wikipedians on this talk page can take advantage of these comments and improve the quality of the article accordingly.

Dr. Hyndman has published scholarly research which seems to be relevant to this Wikipedia article:

  • Reference : Alexander Dokumentov & Rob J Hyndman, 2013. "Two-dimensional smoothing of mortality rates," Monash Econometrics and Business Statistics Working Papers 26/13, Monash University, Department of Econometrics and Business Statistics.

ExpertIdeasBot (talk) 11:14, 1 June 2016 (UTC)