Bayesian Information Criterion Plots and Interpretation Issues

Greetings all,
I am having trouble interpreting the results from an Expected Maximization clustering using mclust and the Iris flower data.

If one were to investigate Plot 1 and 2 ;
How do the graphics lead one to determine the optimum clustering should be 3?


This post has a nice quick explanation (it starts with k-means, but then goes on to mclust)

1 Like