- 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 was delete. -- Cirt (talk) 00:14, 27 October 2010 (UTC)[reply]
- Piecewise regression analysis (edit | talk | history | protect | delete | links | watch | logs | views) – (View log • AfD statistics)
- (Find sources: Google (books · news · scholar · free images · WP refs) · FENS · JSTOR · TWL)
A single-source article where the main author fails to follow Wikipedia guidelines in many respects, despite discussion on Talk page. While the article has multiple references, the topic is about some difference from standard procedures, where there is only a single source for this idea, and where this idea is not actually explained. There are many other problems. My suggestion is Userfy rather than deletion, but there seem many hurdles to cross before this article could be made acceptable. Melcombe (talk) 12:46, 19 October 2010 (UTC)[reply]
As Melcombe said, this article is from a single soruce. Right, since so far there is only one paper in a conference proceedings (Note: please see Correction below ----Yuanfangdelang (talk) 14:05, 25 October 2010 (UTC)). However, the ideas as well as the analytical logic of the new method are right. The current ones take an optimization to determine the threshold as well as the expectation of piecewise models. This is wrong. Why? The segmented variabel X, the combined residuals R of all piecewise models and the matrix M of the piecewise regression coefficients are all random variables in the data iteration for searching the unknown threshold. How can we use the min(R) to determine the expectation of the M, E(M)? Is the correspondence between the min(R) and the E(M) a certain correspondence? The answer is NOT. It is a random correspondence. What corresponds to the E(M) is the E(R) but not the min(R). Thus, the current methods are theoretically incorrect in Mathematics!!![reply]
The second serious mistake is the assumption of the enforced continuity. This is not a statistical hypothesis, no one can assume the continuity in a random space if the unknown threshold(s) exist in it. In a staitstical point of view, we need a probabilistic inference for the continuity.
Since the optimization is wrong, and the continuity cannnot be enforcedly assumed, we cannot take the current methods to estimate the unknown thresholds. We must find another way.
No one has ever doubted the current methods. Ligong Chen is the first and the only one.
I strongly urge you to retain these new ideas so that more and more people could have an opportunity to realize the mistakes in the current system. However if you think that this article violate the wiki rules due to some original researches, I can eliminate the part of "Preperation of Concepts".
I have invited many people to discuss the new analytical logic and the new method. Actually the new method had been discussed with many statisticians at the two conferences (2007 JSM and 2009 JSM), no one could deny it. I believe that more and more people will accept it.
Due to my English writing skill is poor, I wish someone could help it and clean up all linguistic issues.
Anyway, please give me a last notice before you delete this article from the wikipedia if you insist to delete it. I would like to make a copy before the deletion.
Thanks a lot! ----Yuanfangdelang (talk) 04:47, 20 October 2010 (UTC)[reply]
|}
- As I have stated at the discussion of the webpage of this article, I am still working on this article, and once I finish it I will propose to merge it with the segmented regression since both as well as the spline are belong to a same domain. However, in my personal opinion, we need a formal terminology in the area, and the "piecewise regression" and the "threshold" should be the best than any others. I would like to emphasize again, the current methods are theoretically incorrect in Statistics as well as in Mathematics since they violate the fundamentals of the Statistics. ----Yuanfangdelang (talk) 13:48, 20 October 2010 (UTC)[reply]
Evaluate with the Deletion Policy
[edit]According to the weblink, I carefully checked the policy, and made an evaluation for the article as following:
- "Copyright violations and other material violating Wikipedia's non-free content criteria"
The article does not violate any copyright.
- "Vandalism, including inflammatory redirects, pages that exist only to disparage their subject, patent nonsense, or gibberish"
The article is clearly not a vandalism but a serious introduction to a new statistical method.
- "Advertising or other spam without relevant content (but not an article about an advertising-related subject)"
The article is not an advertising or other spam.
- "Content forks (unless a merger or redirect is appropriate)"
Yes, the article can be considered to be merged with the existing article, such as segmented regression and/or Spline,etc.
- "Articles that cannot possibly be attributed to reliable sources, including neologisms, original theories and conclusions, and articles that are themselves hoaxes (but not articles describing notable hoaxes)"
The article cited many literatures from the published journals and conference proceedings. It is not a hoax.
- "Articles for which thorough attempts to find reliable sources to verify them have failed"
The source of the article is reliable since it is an official website of the American Statistical Association, and the paper published in the conference proceedings can be officially cited in any case.
- "Articles whose subjects fail to meet the relevant notability guideline (WP:N, WP:BIO, WP:MUSIC, WP:CORP and so forth)"
I don't exactly understand this term.
- "Articles that breach Wikipedia's policy on biographies of living persons"
The article does not breach the policy on any of living person.
- "Redundant or otherwise useless templates"
The article is not redundant but still in construction and will merge with others.
- "Categories representing overcategorization"
The article does not represent an overcategorization. It is in the domain of Regression in Statistics.
- "Files that are unused, obsolete, or violate the Non-free policy"
The article is not unused or obsolete. It is new. However, I cannot be ensured if it violates the non-free policy on not.
- "Any other use of the article, template, project, or user namespace that is contrary to the established separate policy for that namespace."
The article is not for a personal purpose but for introducing a new method into a wider range of the public.
- "Any other content not suitable for an encyclopedia"
This term needs to be clarified by an authorized person.
Therefore, I strongly urge that you change your mind to delete it from the wikipedia. A formal merge procedure is needed as soon as possible once I finish it. I believe that the people in the future will realize that this article as well as all the discussion on it will be an important event in the history of Statistics since it will cause a strong impact to the current knowledge system of Statistics. ----Yuanfangdelang (talk) 20:34, 20 October 2010 (UTC)[reply]
|}
- May I have your reason for you changed your mind? Please give your discussion in detail but not just a statement. This is not a mathematical style. I have discussed term-by-term to clarify that the article is not subject to the delettion policy of the wikipedia. So, you must give your discussion here. Thanks! ----Yuanfangdelang (talk) 22:52, 20 October 2010 (UTC)[reply]
- Delete - agree with nominator. Lengthy OR essay-style expansion of idea with a single primary source - does not show significant coverage in secondary sources, which is required by WP:NOTE. Articles cannot be justified on the grounds that the topic may become notable in the future. No objection to moving essay to user space if the author wishes to work on it outside article space. Gandalf61 (talk) 09:36, 23 October 2010 (UTC)[reply]
- Comment: This reads more like a thesis than an encyclopedia article. An article could probably be written on Piecewise regression with multiple secondary sources and giving an overview of the subject rather a narrow aspect of it. I'm not sure how much of the material here would be included in such an article but I don't think it would be much. It should be noted that the majority of the edits are from a single user and the majority of that user's edits are in this article. This kind of "article as personal web page" practice should be discouraged. It should also be noted that the principal author has created created a number of redirects to this page, many of which have little to do with the subject.--RDBury (talk) 13:38, 20 October 2010 (UTC)[reply]
Some people even including myself here thought the methodology in the article Piecewise regression analysis came from a single source. This was wrong. It exists at least two different public sources. One is the Chinese journal of Public Health; and the other is the proceedings of the 2007 JSM and the 2009 JSM. Actually if we take into acount all the other literature in the references, the whole methodology came from more than just the two soruces above since the new method is based on the criticisms on the all existing theories and methods. Therefore, it is a multiple-source article; and the method is an improved one. Thus, I cannot agree with the deletion proposal except in the case that thoes who proposed to delete can prove the all sources listed in the references are not reliable public sources; or they can prove that the new methodology as well as its fundametnals are totally wrong. In fact, no bady can do either.----Yuanfangdelang (talk) 14:03, 25 October 2010 (UTC)[reply]
You cannot hide my comments and even delete my correction. Thanks! ----Yuanfangdelang (talk) 21:29, 26 October 2010 (UTC)[reply]
- 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.