We will work with the dataset \(\mathsf{vocabulary}\). 64 observations and
5 variables are available. Let us first apply the K-means algorithm. We
will consider only the first 30 observations for clarity.
We splitted the observations into 6 clusters. From the plot, we can see
by the way that Grade 8 increases with Grade 9. The clusters are chosen
in the way that the variability in the cluster is small. We can read
from the plot the means in each cluster as well.
For comparison, we can use the hierarchical aglomerative clustering algorithm: