K-Prototypes is a combination of K-means and K-modes. This project used K-prototypes, found in the clustMixType package to cluster a data set with 73% numerical variables and 27% of categorical variables.
-
Interpretation of the clusters is done by visualising the clusters using
cprofiles()
. -
Performance comparison of the two clustering algorithms is measured using the rand index
Below is the link to the project implementation https://moloscripts.github.io/Clustering-ICT-Skills/