2
$\begingroup$

The plots show the output of NbClust(). By looking at the plot, is that correct to say that k=5 is the optimal number of clusters?

> nc <- NbClust(m3, distance="euclidean", min.nc=2, max.nc=20, 
+               method="kmeans", index="dindex")
[1] "*** : The D index is a graphical method of determining the number of clusters.  
    In the plot of D index, we seek a significant knee (the significant peak in Dindex 
    second differences plot) that corresponds to a significant increase of the value of 
    the measure."  

"All 10829 observations were used."

enter image description here

Improve this question
$\endgroup$

1 Answer 1

Reset to default
3
$\begingroup$

Unfortunately, all of these heuristics are quite fragile.

In particular, they are usually quite sensitive to data preprocessing, such as normalization.

You may want to try different normalizations, and see if you get similar results, or not...

Also make sure to use visualization - if the clusters don't make sense in the visualization, then the D index probably did not work.

Improve this answer
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.