Wikipedia:Articles for deletion/Optimal classification: Difference between revisions
No edit summary |
|||
Line 37: | Line 37: | ||
*'''Delete and move to Academic Publishing Wiki'''. The problem being described here is merely the problem of decision-tree pruning - which happens to have much simpler and better-documented solutions, and already has an article (although it's in a bad state at the moment). Examples of relevant works are "[http://citeseer.ist.psu.edu/kearns98fast.html A Fast, Bottom-Up Decision Tree Pruning Algorithm with Near-Optimal Generalization]" and "[http://isi.kfupm.edu.sa/journals/pdf/A/an_efficient_algorithm_for_optimal_pruni_almuallim_isi_a1996uu50800006.pdf An Efficient Algorithm For Optimal Pruning Of Decision Trees]". Unsurprisingly, being a biologist, Dr. Rypka merely rediscovered a very simple classical algorithm that has long been known in computer science under a different terminology; his work is not reliable because it has not been peer-reviewed; and it is original research to attempt to generalize this work when no published work has done so. I also think Julie will be content to see their work written up in a place that allows and encourages original research. [[User:Dcoetzee|Dcoetzee]] 18:09, 24 July 2008 (UTC) |
*'''Delete and move to Academic Publishing Wiki'''. The problem being described here is merely the problem of decision-tree pruning - which happens to have much simpler and better-documented solutions, and already has an article (although it's in a bad state at the moment). Examples of relevant works are "[http://citeseer.ist.psu.edu/kearns98fast.html A Fast, Bottom-Up Decision Tree Pruning Algorithm with Near-Optimal Generalization]" and "[http://isi.kfupm.edu.sa/journals/pdf/A/an_efficient_algorithm_for_optimal_pruni_almuallim_isi_a1996uu50800006.pdf An Efficient Algorithm For Optimal Pruning Of Decision Trees]". Unsurprisingly, being a biologist, Dr. Rypka merely rediscovered a very simple classical algorithm that has long been known in computer science under a different terminology; his work is not reliable because it has not been peer-reviewed; and it is original research to attempt to generalize this work when no published work has done so. I also think Julie will be content to see their work written up in a place that allows and encourages original research. [[User:Dcoetzee|Dcoetzee]] 18:09, 24 July 2008 (UTC) |
||
*Also, I must note that the practice of using multiple users who (based on writing above) are obviously the same person to offer an article an air of legitimacy is an abusive practice that must be discontinued at once. This is exactly why we frown on multiple accounts. [[User:Dcoetzee|Dcoetzee]] 18:19, 24 July 2008 (UTC) |
*Also, I must note that the practice of using multiple users who (based on writing above) are obviously the same person to offer an article an air of legitimacy is an abusive practice that must be discontinued at once. This is exactly why we frown on multiple accounts. [[User:Dcoetzee|Dcoetzee]] 18:19, 24 July 2008 (UTC) |
||
:*The article has been moved to the Academic Publishing Wiki. [[Special:Contributions/71.100.167.93|71.100.167.93]] ([[User talk:71.100.167.93|talk]]) 22:43, 24 July 2008 (UTC) |
|||
*'''delete and block socks'''. At the best the subject is not notable. At the worst the article constitutes original research. Article covers the work of two papers supposedly published in the 1970's. I'm not a subject matter expert but I can go out on a limb and say that if an article can't provide clear references, a clear explanation of the topic or evidence of secondary coverage, there is a problem. [[User:Protonk|Protonk]] ([[User talk:Protonk|talk]]) 21:25, 24 July 2008 (UTC) |
*'''delete and block socks'''. At the best the subject is not notable. At the worst the article constitutes original research. Article covers the work of two papers supposedly published in the 1970's. I'm not a subject matter expert but I can go out on a limb and say that if an article can't provide clear references, a clear explanation of the topic or evidence of secondary coverage, there is a problem. [[User:Protonk|Protonk]] ([[User talk:Protonk|talk]]) 21:25, 24 July 2008 (UTC) |
Revision as of 22:43, 24 July 2008
- Optimal classification (edit | talk | history | protect | delete | links | watch | logs | views) (delete) – (View log)
Per Wikipedia:No original research. As a student in the field, I can offer three general observations:
- The article as a whole is unsourced and not verifiable - the "references" deal with statistical classification as a whole and in no way related to article's subject. A Google search will give you many results, but keep in mind that optimal classification is a very general term - one might expect the same with "optimal path planning", but there is no algorithm titled "optimal path planning" (which is what this page is about, an algorithm, as stated in the intro)
- On this point you have indeed shown that you are an unqualified student in this field. Dr. Rypka's work, like so much of modern research is performed at institutes and corporations and only reported briefly due to the need to keep trade secrets of which general academia at universities and schools are simply not privy to or aware of. As for the results of a Google search on the keywords "Optimal Classification" there are some which address nothing of the sort. However, I am quite confident that neither you or anyone else can refute the validity of the article and on that note I invite you to try. What you will end up doing as you have already chosen to do here is to discredit your own validity as a person and as a student. Julie Dancer (talk) 14:15, 24 July 2008 (UTC)
- The equations are not general, well-known equations in the field of pattern recognition. They are individually not verifiable in the context of classification.
- As for the merit of the equations... In the course of studying them myself for the purpose of providing variable labels and descriptions I was quite delighted to find the opposite condition to be true, that the components are used frequently enough in various areas that comparisons of their use may have been the only thing which made detailed description of the variables possible. But again you are challenging the validity of the equations on the basis of your student opinion rather than any valid example, demonstration or work. Julie Dancer (talk) 14:30, 24 July 2008 (UTC)
- The single example provided, the "Flag Recognition - Application Example" section, was a modified version of this, and is an example of classification with neural networks in general, and cannot provide verification for this article.
- The example is not intended to provided verification but only example of usage. Julie Dancer (talk) 14:35, 24 July 2008 (UTC)
In addition, I'd like to direct your attention to the contribution history of the article - aside from minor contributions from a small set of editors, I believe the following users:
Are the aliases of User:Pce3@ij.net, who was blocked indefinitely for contributing original research articles (see Wikipedia:Articles for deletion/Rapid sort). Since then, Pce3 republished his ideas outside of Wikipedia, on Wikia here, which can be used to verify that the three accounts are linked. Without addressing Wikipedia:Sock puppetry, the Wikia articles in particular bring doubt over the credibility of the editor.-- Jiuguang (talk) 10:55, 24 July 2008 (UTC)
Comment
- Optimal Classification was developed for the purpose of identifying microbes using the least number of queries. It performs a task which many people perform everyday and some pride themselves in doing better than others. However, the reason the process is named "optimal" is because there may exist more than one sequence of queries which are minimal. Thus "optimal" is the preferred terminology to "minimal." Like the lack of affection many aspiring chess players have for IBM's Deep Blue such a technique is not relished by players of the game 20 questions.
- To avoid jealousy and controversy, I personally have no objection if the article is deleted here and moved to the Wikia since it is a great imposition to check up on it here and to be sure that vandalism has not occurred or changes made which would render it less accurate or clear. However, my own belief is that this request for deletion has nothing to do with the criticism which has been stated but rather with personal academic jealousy and ignorance on the part of Jiuguang rather than any lack of credibility or merit on the part of the subject matter. Julie Dancer (talk) 13:20, 24 July 2008 (UTC)
- As for Jiuguang's accusation of the article being original research... The topic, like so many, which I find interesting is not necessarily mainstream, just as it is hard to find municipalities which convert their solid organic waste and bimass into biochar, bio-oil and syngas at this time versus spreading it over landfills or burning it at high temperatures in some cases to generate power but with far greater waste as an all around resource. My contribution to the article lies mainly in developing a clear and concise example and detailing the variables involved in the equations to help readers to better understand what Optimal Classification is all about. The original research of the topic matter was done by Dr. Rypka, et. el. But again, I have no objection to deletion of the article and its publication somewhere else. Julie Dancer (talk) 13:20, 24 July 2008 (UTC)
- Aside from the personal attacks - Eugene W. Rypka published no algorithm titled "optimal classification" (please provide your source?) Rypka works in classification and identification in general. --Jiuguang (talk) 14:13, 24 July 2008 (UTC)
- optimal and optimal —Preceding unsigned comment added by Adaptron (talk • contribs) 15:22, 24 July 2008 (UTC)
- If you do not want to subject yourself to being discredited then I suggest you not declare yourself to be expert in the field as grounds for deletion of an article that is over your head.
- No, actually Dr. Rypka worked before his death in the field of microbiological identification to which he devoted his work in pattern recognition. His work, however, is so universal that it need not be limited to microbial identification, since he provided the principles in general and published them in that vein. Julie Dancer (talk) 14:46, 24 July 2008 (UTC)
- You do not have my permission to move what I write away for the topic it is addressing until I am finished completing the thought. Julie Dancer (talk) 14:32, 24 July 2008 (UTC)
- Aside from the personal attacks - Eugene W. Rypka published no algorithm titled "optimal classification" (please provide your source?) Rypka works in classification and identification in general. --Jiuguang (talk) 14:13, 24 July 2008 (UTC)
- Delete—First off I am supremely unqualified to judge this article on its subject matter. But judging soley on the policies of Wikipedia, the article—in its current form—doesn't meet the standards of an article here. The "references" in the article are currently impossible to decypher, and there are no citations in the article text to give a clue what the references are supposed to be supporting. The suggestion that the article is original research stems from this lack of clear referencing. If this is a subject that has verifiable sources that actually apply to the topic of this article, then by all means add them to the article and I'll change my mind. Livitup (talk) 16:14, 24 July 2008 (UTC)
- Actually I hope you do not change your mind since I have done enough work on the article. The effort here is suppose to be collaborative but instead it appears that even if someone else was knowledgeable about this topic they would be too lazy to contribute to it anyway. In fact, I am in the process of finding a new home for it right now. Something perhaps on the order of this: "It's been seven months since Google (Nasdaq: GOOG) introduced Knol, a Wikipedia knockoff that raises the stakes by offering original authors opportunities for content control, acclaim, and cold, hard cash." - from: The Motley Fool upon consensus. -- adaptron (talk) 16:35, 24 July 2008 (UTC)
- Another thing of course is that at one time I thought verifiable sources meant verifiable content. I was all enthused that what I read here was certified as valid whereas now that I realize that verifiable sources means only verification that something has been printed, put on a library shelf and indexed so that at least 50% of the actual knowledge in the world may not even be trade secrets out there intentionally withheld from publication. This not only leaves a big, big gap in what is possible to find here but that the validity of the content is based only on the fact that the binding can be found on a shelf. A sad disappointment indeed. However, I have republished the article now in no less the four places so I can be out of here. -- adaptron (talk) 17:23, 24 July 2008 (UTC)
- Delete and move to Academic Publishing Wiki. The problem being described here is merely the problem of decision-tree pruning - which happens to have much simpler and better-documented solutions, and already has an article (although it's in a bad state at the moment). Examples of relevant works are "A Fast, Bottom-Up Decision Tree Pruning Algorithm with Near-Optimal Generalization" and "An Efficient Algorithm For Optimal Pruning Of Decision Trees". Unsurprisingly, being a biologist, Dr. Rypka merely rediscovered a very simple classical algorithm that has long been known in computer science under a different terminology; his work is not reliable because it has not been peer-reviewed; and it is original research to attempt to generalize this work when no published work has done so. I also think Julie will be content to see their work written up in a place that allows and encourages original research. Dcoetzee 18:09, 24 July 2008 (UTC)
- Also, I must note that the practice of using multiple users who (based on writing above) are obviously the same person to offer an article an air of legitimacy is an abusive practice that must be discontinued at once. This is exactly why we frown on multiple accounts. Dcoetzee 18:19, 24 July 2008 (UTC)
- The article has been moved to the Academic Publishing Wiki. 71.100.167.93 (talk) 22:43, 24 July 2008 (UTC)
- delete and block socks. At the best the subject is not notable. At the worst the article constitutes original research. Article covers the work of two papers supposedly published in the 1970's. I'm not a subject matter expert but I can go out on a limb and say that if an article can't provide clear references, a clear explanation of the topic or evidence of secondary coverage, there is a problem. Protonk (talk) 21:25, 24 July 2008 (UTC)