Talk:Classification

Proposed major reorganisation
This page reads like it has been written from one, very particular academic viewpoint, presumably the school concerned with "classification theory", the page that was merged into here. It almost completely fails to address, and when it does mention drastically downplays, the very substantial treatment of classification in subjects such as law, mathematics, statistics, economics, machine learning, cognition, taxonomy, anthropology, philosophy and sociology. This is I think the explanation for why the page is at the same time quite extensive and considered vitally important by Wikipedia, but at the same time rated low in quality.

I therefore propose to introduce brief, referenced sections on all these treatments of classification, to relegate the existing page to a single section at the same level, and to introduce a much more general and approachable introduction.

Untitled
Classifications, is the act of sorting objects/items, etc. according to one or more characteristics of the latter into groups or sets. It was highly desireable that the members of the resulting class shall be selected by using normally one criterion and consistently all the way through, otherwise your class will not be homogenous or fully sorted(another desireable feature). It is also often required that on classifying a given set the criteria to be used for clasifying things should allow for a full allocation of the input set to appropriate classes. The list of classified objects arranged accordingly will comprise a classification or nomenclature that play an important role in distributing/sharing knowledge about products, trade marks, and thousands of other whatnots in the world. apogr


 * Oh. Actually I was hoping to find some explanation here on the talk page. As I see it, classification is an everyday activity and not just something for librarians, biologists and machine learners. For example, there are hotel classification schemes such as the one to five stars. Shouldn't we add this somehow here?  05:19, 16 Jul 2004 (UTC)


 * I do not agree with the current treatment of articles as examplified in the above case too for the following reasons:

the introduction of a subject is done by introducing sometimes two or three different and usually unknown, or two wide subjects to garsp as a result of which one cannot focus, but must progress on a wide mental horizon. Then by partially giving account of the uses/senses of the word or by quasi disambiguation, again a similar reasult/depth is likely. What I miss here is connecting the unknown to the known and exposig the point why the article may be of interest to the general reader. If you vet the current article against my comments above, I hope you will see the point I am trying to make.

Many wikipedia articles are just the names of boxes without specifying knowledge/content, except for the labels of smaller or bigger boxes inside, nested just one or two levels. IMHO such a beautiful venture to become a hit for the reader too, should have a feature like highlighting the current problems/prospects/controversies of the subject covered in the article so that the reader can have something to nible on.

You will have remembered that information is specific and not general and you shuld visibly and aptly move along the specific-generic continuum to become educative/informative.

Apogr 06:31, 20 Aug 2004 (UTC)

Medical classification
I think that Medical classifications can be both - scientific and statistical. The examples given in Medical classification are to a large. So medical classifications like ICD are definitely examples for statistical classifications - and this is what I wanted to express. Nosology on the other hand is an example for a scientific medical classification.

Might be that there is an additional meaning of "statistical classification" and ICD doesn't fit within the list of examples. The examples given are examples for algorithms building up a statistical classification. The ICD is a statistical classification (that was not build using algorithms but manpower). It is used for making statistics and such is a statistical one. If you look at Ontology (computer science) and compare the definition "an ontology is the attempt to formulate an exhaustive and rigorous conceptual schema within a given domain" then for example ICD would be a bad example.

I propose to make two sub-section of statistical classification. The first contains the current definition and the second has the definition "A statistical classification is a classification made for statistical purposes. It categorizes real world things (objects, events, ...) into classes of things with common properties."


 * Any human-built classification scheme can be used for statistical purposes, so I don't see how that would be lead to a distinction --- for example, people compute statistics over scientific classifications, to see how active different fields are. Why can't people compute statistics over Nosology? In the field of statistics, classification is a type of algorithm, not an human-built ontology. The proposed meaning of "statistical classification" is not in common usage and would completely blur the two distinct meanings. -- hike395 08:22, 20 Oct 2004 (UTC)


 * Yes, you are right and on the other hand you're not right, since you have to deal with language use. In fact, the names of some classifications contain the adjective "statistical", e.g. "International Statistical Classification of Diseases and Related Health Problems" or "Diagnostic and Statistical Manual of Mental Disorders". And the main intention of some specific classifications is to make statistic analyses. I think that taxonomies and ontologies don't suit to this intention. The first section in the main part of this talk is a good defintion of what I mean. If you interprete ontology rather widely as more or less any system where something can be assigned to something like classes then the mentioned medical classification are ontological classifications - still having the "statistical" adjective. --Udo Altmann 09:48, 20 Oct 2004 (UTC)


 * Sounds like you're objecting to the use of the term ontology, rather than the underlying distinction between man-made vs. AI. Let me take a crack at changing the first part, to see if I can match your definition. -- hike395 14:31, 20 Oct 2004 (UTC)


 * I agree. Thank you. Confusingly I can see the text only in the history, not yet directly under http://en.wikipedia.org/wiki/Classification - is there a feature that I don't know? (It's not the refresh button ... and I'm not aware of any proxy ...) --Udo Altmann 15:05, 20 Oct 2004 (UTC)


 * Wikipedia has been flaky (I keep getting logged off, for example). There is probably a cache on the server. Give it a try now, I see it. -- hike395

Security classification
I think that security classification is a special kind of taxonomic classification and shut be put as an additinal examle there: "Taxonomy may refer to either a hierarchical classification of things ...". In daily life we have a lot of classifications of such kind: hotel (see above), credit rating ... --Udo Altmann 08:30, 12 Jan 2005 (UTC)

taxonomic vs. statistical classification
Hi, why is it that we classify classification like that? The method, process, and result of a statistical classification, as defined in the article, is an instance of taxonomical classification, or is it not? If not, why not? :-) --Glimz &shy;, 04:08, Jan 24, 2005 (UTC)


 * Yes, once you train up a statistical classifier, if it is accurate, it can replace a human and place objects in a taxonomy automatically. I think we group the two different classifications differently, because some people really just want to look up taxonomies, while other people are interested in automatic classification algorithms. This started out as a disambiguation page --- it just sort of grew into a stubby article. -- hike395 06:47, 24 Jan 2005 (UTC)

Definitions to Classification = To classify something or something to get classified, classifying something , or too use a classifier.

classification of mushroom - by mohit
classification of mushroom is should be ask to amol sir - 9321712857 —Preceding unsigned comment added by 117.204.161.209 (talk) 12:54, 26 February 2010 (UTC)

Induction & deduction
WP DEPENDS ON INDUCTION & DEDUCTION. THEIR MEANING SHOULD BE FIRM. This encyclopedia accepts the premise of enumerative induction that the more editors who agree on the content of an article, the more accurate and useful that content. Induction is practiced on every TALK page. Editors generalize from a few observations, and deduce concrete conclusions from their generalizations.

WP contains 4 repetitive and fragmentary articles on induction: [Inductive reasoning], [The problem of induction]; [New riddle of induction],[Inductivism]. I would like to rectify this chaotic situation by rewriting and merging these 4 articles, retaining only the reasoning title. I ask you—a participant in relevant TALK pages—to judge my rewrite/merge project: SHOULD I PROCEED? Below is the current proposed outline:

Definitions. Induction generalizes conceptually; deduction concludes empirically.

[David Hume], philosopher condemner.

[Pierre Duhem], physicist user.

[John Dewey], philosopher explainer.

[Bertrand Russell], philosopher condemner.

[Karl Popper], philosopher condemner.

Steven Sloman, psychologist explainer.

Lyle E. Bourne, Jr., psychologist user.

[Daniel Kahneman], psychologist user.

[Richard H. Thaler] economist user.

Please respond at Talk:Inductive reasoning. TBR-qed (talk) 16:10, 5 January 2020 (UTC)