Understand linkage in dendrogram for regrouping original data points

2 次查看（过去 30 天）
Florian
Florian 2018-12-18
Hi all,
I have a 516 x 4 matrix X representing 516 data points with 4 observations each. Using hierarchical clustering I group the data points according their relation. I follow the following steps:
T=clusterdata(X,'distance','squaredeuclidean',1);
tree = linkage(T,'ward');
figure, dendrogram(tree)
The first operation creates an array T of size 516 x 1 which assigns a number to each data point including identical numbers for similar data points. Here it is still easy to associate the original data with the cluster results.
Next, from the linkage I get 515 x 3 matrix tree. This matrix has three rows, while the subsequent dendrogram has linkages at several levels within the dendrogram (see attached).
How can I find out from the matrix tree which original data points are lumped together into the various branches of my dendrogram?
Best regards,
Florian
(Matlab 2018b with Statistics & Machine learning toolbox)

请先登录，再回答此问题。

回答（0 个）

类别

在 Help Center 和 File Exchange 中查找有关 Hierarchical Clustering 的更多信息

标签

产品


版本

R2018b

Community Treasure Hunt

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

Start Hunting!

Translated by