Find a best k for the k-means clustering

Find a best k for the k-means clustering

find_best_km(mat, max_km = 15)

Arguments

mat

A matrix where k-means clustering is executed by rows.

max_km

Maximal k to try.

Details

The best k is determined by looking for the knee/elbow of the WSS curve (within-cluster sum of square).

Note this function is only for a rough and quick estimation of the best k.

Examples

# There is no example
NULL
#> NULL