I have plant biomass and grain weight data for different plant varieties which I now need to cluster. Do I need to define the number of clusters if using Ward's method and Squared Euclidean distance in SPSS? Also both of the variables are measured in same units so do I need to standardize this data?
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
1
Ward's method is a hierarchical clustering procedure, so number of cluster is not required to start (unlike k-means procedure, which requires number of cluster to be given beforehand).
Standardization is not required if the two dimensions are roughly on the same magnitude.
-
$\begingroup$ Thanks @Nuclear03020704, another doubt is whether there is any co-efficient associated to each cases defining how well they are classified in each cluster? If not how can we understand how good/bad are the clusters? $\endgroup$ Jan 25, 2021 at 4:32