I am working with a dataset of volumes that I have classified into components using the mclust package in R. (univariate, unequal variance).
I am now looking for the best way to determine cutoff values (volumes) between each component.
The way I can see to do this is to use the density estimations of each population and find the intersection point between two adjacent populations that would then serve as the cutoff value. However, these intersections do not correspond to the classification and uncertainty variables given to me by mclust.
Am I better off just using the classification and uncertainty from mclust? is there a reason why the density estimations don't gave me the same cutoff values used by mclust?
Thanks in advance!