Jump to content

Talk:Multivariate adaptive regression spline

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

Overall style

[edit]

This article is quite impressive. However, it reads too much like a how-to manual or a textbook. I may have a go to clean that up. Baccyak4H (Yak!) 16:54, 22 July 2008 (UTC)[reply]

I just read this article - and I found it extremely well written and clear. One of the best examples of why wikipedia rocks. Thanks!--Achristoffersen (talk) 22:41, 13 May 2009 (UTC)[reply]
I agree that it reads like a manual or textbook, but I think that is the reason it is an accessible explanation and a great Wikipedia article. Other explanations I have read on blogs or in textbooks is not as clear as this one. I believe it should be left as is. Fairybluebirb (talk) 09:40, 18 April 2023 (UTC)[reply]

Personally, I think it's a great article, and wouldn't change a thing. I expected to have to dedicate at least an hour of jumping back and forward between definitions in other articles, but I walk away after 20 minutes with a good grasp of the subject. It guided me through clear, concise examples, with an accessible but not oversimplified explanation, and I found the tone a pleasure to read. It reads more like a seasoned teacher than like a scientific article (of which I've had my fill these last few days), without sacrificing seriousness. Thank you! 213.245.23.234 (talk) 19:24, 22 July 2013 (UTC)[reply]

[edit]

@2604:2000:E083:9400:AC90:4488:F6B8:1E6D: The article is not in violation of copyright:

The Copyvio Report reports similarities to https://www.slideshare.net/Eklavyagupta/multivariate-adaptive-regression-splines-71969518 . That web page was published in Feb 2017, as stated on the web page.

The original text for this Wikipedia article "Multivariate adaptive regression splines" was published in July 2008.

By comparing these dates, we see that the Slideshare web page mentioned above copied the Wikipedia article, not the other way round.

Stephen Milborrow (talk) 02:33, 15 June 2017 (UTC)[reply]

Indeed; similar considerations apply to the source mentioned by the IP, Jeffrey Strickland (2014), Predictive Analytics using R, Lulu Inc. I've left a {{backwardscopy}} tag for that one. Justlettersandnumbers (talk) 18:53, 7 July 2017 (UTC)[reply]

Error in the example formula?

[edit]

I think the formula is missing some terms just before max(0, 200 - vis). The formula should be [...] 0.016 max(0, wind - 7) + c max(0, 200 - vis) where c is a small coefficient (about 0.035 if I had to guess). TyrHexFF (talk) 08:35, 15 March 2023 (UTC)[reply]

My bad I was wrong. TyrHexFF (talk) 08:37, 15 March 2023 (UTC)[reply]
[edit]
Further information: Wikipedia: External links
Some things just grow during incremental edits and sometimes get out of hand. The "External links" section, one of the optional appendices, was expanded to 13 entries, organized into four subsections is excessive. Three seems to be an acceptable number, and of course, everyone has their favorite to try to add for a fourth. Consensus needs to determine this. A tag indicates concerns.
However, none is needed for article promotion.
Some links may be included in WP:ELNO, or What Wikipedia is not (policy) such as WP:NOTREPOSITORY or WP:NOTGUIDE.
  • WP:ELDEAD may apply.
  • In some cases ELCITE applies: Do not use {{cite web}} or other citation templates in the External links section. Citation templates are permitted in the Further reading section. Others, listed below:
  • ELpoints #3) states: Links in the "External links" section should be kept to a minimum. A lack of external links or a small number of external links is not a reason to add external links.
  • LINKFARM states: There is nothing wrong with adding one or more useful content-relevant links to the external links section of an article; however, excessive lists can dwarf articles and detract from the purpose of Wikipedia. On articles about topics with many fansites, for example, including a link to one major fansite may be appropriate.
  • ELMIN: Minimize the number of links.
The External links guideline This page in a nutshell: External links in an article can be helpful to the reader, but they should be kept minimal, meritable, and directly relevant to the article. With rare exceptions, external links should not be used in the body of an article.
Second paragraph, acceptable external links include those that contain further research that is accurate and on-topic, information that could not be added to the article for reasons such as copyright or amount of detail, or other meaningful, relevant content that is not suitable for inclusion in an article for reasons unrelated to its accuracy.
    • Please also note:
  • WP:ELBURDEN: Disputed links should be excluded by default unless and until there is a consensus to include them. Please do not add back more links without consensus. Simple solution to facilitate career maintenance tag. Move links here for discussion.
Moved links:

Several free and commercial software packages are available for fitting MARS-type models.

Free software
Commercial software
Also move unsourced material and remove 2016 career maintenance tags:

No regression modeling technique is best for all situations. The guidelines below are intended to give an idea of the pros and cons of MARS, but there will be exceptions to the guidelines. It is useful to compare MARS to recursive partitioning and this is done below. (Recursive partitioning is also commonly called regression trees, decision trees, or CART; see the recursive partitioning article for details). 

Compare the equation for ozone concentration above to, say, the innards of a trained neural network or a random forest.
  • MARS tends to be better than recursive partitioning for numeric data because hinges are more appropriate for numeric variables than the piecewise constant segmentation used by recursive partitioning.
The hinge functions automatically partition the input data, so the effect of outliers is contained. In this respect MARS is similar to recursive partitioning which also partitions the data into disjoint regions, although using a different method.
  • MARS models tend to have a good bias-variance trade-off. The models are flexible enough to model non-linearity and variable interactions (thus MARS models have fairly low bias), yet the constrained form of MARS basis functions prevents too much flexibility (thus MARS models have fairly low variance).
  • MARS is suitable for handling large datasets, and implementations run very quickly. However, recursive partitioning can be faster than MARS
  • With MARS models, as with any non-parametric regression, parameter confidence intervals and other checks on the model cannot be calculated directly (unlike linear regression models). Cross-validation and related techniques must be used for validating the model instead.
  • The earth, mda, and polspline implementations do not allow missing values in predictors, but free implementations of regression trees (such as rpart and party) do allow missing values using a technique called surrogate splits.
  • MARS models can make predictions very quickly, as they only require evaluating a linear function of the predictors.
  • The resulting fitted function is continuous, unlike recursive partitioning, which can give a more realistic model in some situations. (However, the model is not smooth or differentiable). -- Otr500 (talk) 08:58, 1 July 2025 (UTC)[reply]
Retrieved from "https://en.wikipedia.org/w/index.php?title=Talk:Multivariate_adaptive_regression_spline&oldid=1298243241"