question about k-means clustering metric choice

5 次查看（过去 30 天）
cgo
cgo 2018-8-17
评论： Walter Roberson 2018-10-3
In this documentation of K-means clustering: https://www.mathworks.com/help/stats/k-means-clustering.html
I would like to ask why the choice of metric is 'cityblock'. May I know what the context of the dataset is and why 'cityblock' is the appropriate choice of metric? Also, I tried to use the ordinary euclidean metric, but returned an error.
Please send your insights. Thanks
  2 个评论
Walter Roberson
Walter Roberson 2018-8-17
编辑：Walter Roberson 2018-8-17
I do not encounter any errors if I consistently change cityblock to sqeuclidean
https://www.mathworks.com/help/stats/sample-data-sets.html?searchHighlight=kmeansdata just describes it as 4 dimensional clustered data.
cgo
cgo 2018-8-17
found it. I typed 'euclidean' instead of 'sqeuclidean'. Thanks

请先登录，再进行评论。

请先登录，再回答此问题。

回答（1 个）

Bernhard Suhm
Bernhard Suhm 2018-10-3
If you have any further questions, follow up. Otherwise, can you kindly "accept" Walter's answer?

类别

在 Help Center 和 File Exchange 中查找有关 k-Means and k-Medoids Clustering 的更多信息

标签

产品

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by