![]() | This article is rated Start-class on Wikipedia's
content assessment scale. It is of interest to the following WikiProjects: | ||||||||||
|
![]() | It is requested that a computing diagram or diagrams be
included in this article to
improve its quality. Specific illustrations, plots or diagrams can be requested at the
Graphic Lab. For more information, refer to discussion on this page and/or the listing at Wikipedia:Requested images. |
Why is this article nominated for deletion? I did not write this but it seem correctly cited and of some marginal worth. The Wikipedia CS and stats pages are already kinda bare bones. It seems arbitrarily spiteful to the author to remove this page when it is of some use and is being removed simply for “not being of note”. Madjack74 ( talk) 15:51, 19 May 2009 (UTC)
This is my first time editing Wiki, so I'm not sure if this is where to discuss, but is the term "vanilla k-means" really a good way to refer to the original algorithm? "Original k-means" sounds much more professional and less hipster-doofus to me... MetaphysicalBrian ( talk) 16:57, 4 June 2009 (UTC)
I think there is a mistake in the "Example Bad Case". Suppose the four points are (0,0), (3,0), (0,8), (3,8), and the initial arbitrary partition is {(0,0), (0,8)}, {(3,0), (3,8)}. In this case the centroids are (0,4) and (3,4), and the total cost is 4 x 4^2 = 64. Without loss of generality suppose we pick point (0,0) and try to improve by moving it to the second cluster, then we get: {(0,8)}, {(0,0), (3,0), (3,8)} In this case the centroids are (0,8) and (2,8/3), and the total cost is 0+11.11+8.11+29.44 = 48.66, which is already better. (Moreover, I didn't check but I think that the next step will move point (3,8) to the first cluster which gives optimal solution). Hagai Cibulski —Preceding unsigned comment added by Hagai Cibulski ( talk • contribs) 21:30, 7 December 2009 (UTC)
I would like to bring up the case for deletion again. Even as an important method for initialisation I fear it is not sufficiently distinct from other forms of initialisation to warrant its own page. Putting a positive spin on things, it would be of more use alongside other forms of initialisation; the discussion on the main K-means page is short. K-means++ should be on Wikipedia in an article alongside the other 101 methods of parameter initialisation, where it will be of more use to readers. SymplectoJim ( talk) 11:24, 22 March 2019 (UTC)
Especially step 3. Can anyone figure out what that is supposed to mean? Seriously, xk+1 > D(xk)2 is just nonsense. In the paper I've found it says something completely different. You're supposed to take as a new center any data point x with probability proportional to D(x)2. So basically, the standard k-means initialization chooses centers randomly among data point, each being equally weighted (or weighted by D(x)0), while k-means++ weights them by D(x)2. 77.23.64.247 ( talk) 13:43, 4 January 2010 (UTC)
Hello fellow Wikipedians,
I have just modified one external link on K-means++. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
{{
dead link}}
tag to
http://www.cse.ohio-state.edu/~johansek/clustering.pdfWhen you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.
This message was posted before February 2018.
After February 2018, "External links modified" talk page sections are no longer generated or monitored by InternetArchiveBot. No special action is required regarding these talk page notices, other than
regular verification using the archive tool instructions below. Editors
have permission to delete these "External links modified" talk page sections if they want to de-clutter talk pages, but see the
RfC before doing mass systematic removals. This message is updated dynamically through the template {{
source check}}
(last update: 5 June 2024).
Cheers.— InternetArchiveBot ( Report bug) 20:13, 3 December 2017 (UTC)
![]() | This article is rated Start-class on Wikipedia's
content assessment scale. It is of interest to the following WikiProjects: | ||||||||||
|
![]() | It is requested that a computing diagram or diagrams be
included in this article to
improve its quality. Specific illustrations, plots or diagrams can be requested at the
Graphic Lab. For more information, refer to discussion on this page and/or the listing at Wikipedia:Requested images. |
Why is this article nominated for deletion? I did not write this but it seem correctly cited and of some marginal worth. The Wikipedia CS and stats pages are already kinda bare bones. It seems arbitrarily spiteful to the author to remove this page when it is of some use and is being removed simply for “not being of note”. Madjack74 ( talk) 15:51, 19 May 2009 (UTC)
This is my first time editing Wiki, so I'm not sure if this is where to discuss, but is the term "vanilla k-means" really a good way to refer to the original algorithm? "Original k-means" sounds much more professional and less hipster-doofus to me... MetaphysicalBrian ( talk) 16:57, 4 June 2009 (UTC)
I think there is a mistake in the "Example Bad Case". Suppose the four points are (0,0), (3,0), (0,8), (3,8), and the initial arbitrary partition is {(0,0), (0,8)}, {(3,0), (3,8)}. In this case the centroids are (0,4) and (3,4), and the total cost is 4 x 4^2 = 64. Without loss of generality suppose we pick point (0,0) and try to improve by moving it to the second cluster, then we get: {(0,8)}, {(0,0), (3,0), (3,8)} In this case the centroids are (0,8) and (2,8/3), and the total cost is 0+11.11+8.11+29.44 = 48.66, which is already better. (Moreover, I didn't check but I think that the next step will move point (3,8) to the first cluster which gives optimal solution). Hagai Cibulski —Preceding unsigned comment added by Hagai Cibulski ( talk • contribs) 21:30, 7 December 2009 (UTC)
I would like to bring up the case for deletion again. Even as an important method for initialisation I fear it is not sufficiently distinct from other forms of initialisation to warrant its own page. Putting a positive spin on things, it would be of more use alongside other forms of initialisation; the discussion on the main K-means page is short. K-means++ should be on Wikipedia in an article alongside the other 101 methods of parameter initialisation, where it will be of more use to readers. SymplectoJim ( talk) 11:24, 22 March 2019 (UTC)
Especially step 3. Can anyone figure out what that is supposed to mean? Seriously, xk+1 > D(xk)2 is just nonsense. In the paper I've found it says something completely different. You're supposed to take as a new center any data point x with probability proportional to D(x)2. So basically, the standard k-means initialization chooses centers randomly among data point, each being equally weighted (or weighted by D(x)0), while k-means++ weights them by D(x)2. 77.23.64.247 ( talk) 13:43, 4 January 2010 (UTC)
Hello fellow Wikipedians,
I have just modified one external link on K-means++. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
{{
dead link}}
tag to
http://www.cse.ohio-state.edu/~johansek/clustering.pdfWhen you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.
This message was posted before February 2018.
After February 2018, "External links modified" talk page sections are no longer generated or monitored by InternetArchiveBot. No special action is required regarding these talk page notices, other than
regular verification using the archive tool instructions below. Editors
have permission to delete these "External links modified" talk page sections if they want to de-clutter talk pages, but see the
RfC before doing mass systematic removals. This message is updated dynamically through the template {{
source check}}
(last update: 5 June 2024).
Cheers.— InternetArchiveBot ( Report bug) 20:13, 3 December 2017 (UTC)