. As implemented by R, K-means with centers = number of clusters randomly selects rows of the data matrix as initial cluster centers. Test the robustness of K-means in R for the scaled hominoid data...

.
As implemented by R, K-means with centers = number of

clusters randomly selects rows of the data matrix as initial cluster centers. Test

the robustness of K-means in R for the scaled hominoid data in Section 10.5

by repeatedly performing the clustering while specifying 2 for centers.What

is there about this data set that produces two different pairs of clusters for

k = 2? Which result is “better”? What do you conclude about the robustness

of K-means with this method of specifying initial cluster centers?