The major difficulty of K-Means is the pre-requisite of the number of the cluster (K) that must be defined before the algor

[answerhappygod]

The major difficulty of K-Means is the pre-requisite of the number
of the cluster (K) that must be defined before the algorithm is
applied to the input dataset.
If the plot results show the centroids are to close to each
other, what should the researcher do first?
- Just reach the conclusion that the given input dataset is not
suitable for this clustering approach.
- Do nothing and analyze the results as it is.
- Do not run K-Means and choose another clustering algorithm
such as the hierarchical one.
-Decrease the number of clusters (K) and re-run the algorithm
again.
-Increase the number of clusters (K) and re-run the algorithm
again.