如何从分层聚类中解释树状图以找到最佳聚类数？
[英]How to interpret a Dendrogram from hierarchical clustering to find optimal number of clusters?
问题描述
When viewing this, how do we know to find the optimal number of clusters? 
查看此内容时，我们如何知道找到最佳数量的群集？ I used K-means and found the "elbow" on the graph that showed the optimal point but I am having trouble figuring this out from just the dendrogram. 我使用了K均值，并在图形上找到了显示最佳点的“肘部”，但是我很难从树形图中找出这一点。
1 个解决方案
解决方案1
0 2017-11-05 08:58:04
The interpretation varies depending on your metric and linkage used. 解释因您的指标和使用的链接而异。
But in general, you want to keep branches that have "many" observations and with a “large" distance above (for the next merge). 但是总的来说，您希望保留具有“许多”观察值且其上方具有“大”距离的分支（用于下一个合并）。
声明:本站的技术帖子网页，遵循CC BY-SA 4.0协议，如果您需要转载，请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.