Talk:Active learning (machine learning)

This article was the subject of a Wiki Education Foundation-supported course assignment, between 12 January 2022 and 4 May 2022. Further details are available on the course page. Student editor(s): Jackstevns (article contribs).

Content is biased towards protein engineering

Untitled

Active learning is a well studied area of machine learning, as the linked survey article by Settles shows. Nontheless, the other references, and much of the article content, relates to two very specific articles regarding active learning in protein engineering. "Maximum curiosity" does not seem to be a particularly well-known example of active learning; it is mentioned without citation in the second Dansiger reference.

Interestingly, this active learning text was largely moved over from the Supervised learning article, in which it was mostly added in 2008 by User:Sad1225. This is likely S.A. Dansiger, the primary author of the protein engineering articles linked. It's fabulous that Dansiger wanted to add some content about active learning to Wikipedia. As it stands, though, this article is a lousy overview of active learning, and portrays a completely misleading sense of the field.

I propose that the content on maximum curiosity and protein engineering should be deleted. WorldsApart (talk) 01:41, 23 May 2012 (UTC)

I made a number of changes to make the article more general. WorldsApart (talk) 11:25, 3 July 2012 (UTC)

Wrong language link

This article must link to https://ja.wikipedia.org/wiki/能動学習 in Japanese wikipedia, not https://ja.wikipedia.org/wiki/プールベース能動学習. I tried to edit myself, but failed because プールベース能動学習 is already linked here and it is forbidden to remove a link. Only renaming is allowed. Please anyone who have rights, fix these links. Singleheart (talk) 01:40, 4 July 2016 (UTC)

Query, pass, iteration, stage, phase, round, or cetera? Or just size?

Often it is the case that one or more examples will be sent to the oracle for labeling, the new label information will be used for training to update the model, and the process will repeat. Is each of these repeat units typically called a "pass", an "iteration", a "stage", a "phase", a "round", or something else? (I am hopeful that it isn't "batch" or "epoch" as they are already meaningful!) The article currently uses "single-pass" for online learning and elsewhere uses "iteration". However, if there is a standard then we should discuss that in the article and we should use it in the article. —Quantling (talk | contribs) 13:19, 13 October 2022 (UTC)

Or ... is it the case that the number of these repeat units isn't usually counted and instead the cumulative size of the dataset is instead reported? —Quantling (talk | contribs) 13:44, 13 October 2022 (UTC)

One more possibility; I am seeing "query" in the literature. This makes sense because it is common to say that "the active learning strategy queries an oracle". —Quantling (talk | contribs) 16:25, 26 October 2022 (UTC)

Wiki Education assignment: Linguistics in the Digital Age

This article was the subject of a Wiki Education Foundation-supported course assignment, between 21 August 2023 and 11 December 2023. Further details are available on the course page. Student editor(s): Jfernandez212 (article contribs).

— Assignment last updated by Fedfed2 (talk) 00:53, 9 December 2023 (UTC)

Wiki Education assignment: Linguistics in the Digital Age

This article is currently the subject of a Wiki Education Foundation-supported course assignment, between 15 January 2024 and 8 May 2024. Further details are available on the course page. Student editor(s): Minhngo6 (article contribs).

— Assignment last updated by Cakelover44 (talk) 03:36, 19 February 2024 (UTC)