For use in the example sentence:
After the outliers had been removed, the ... were grouped with a disjoint set (aka. Union-Find).
I lean towards inliers, but this seems wrong given the linked definition.
"... the remainder were grouped..."
You also might want to use cluster as a term for the grouping too.
TBQH, given that it is 2018 and statistical discussions are more commonplace (people were not discussing Gladwell or Piketty when they spoke Latin), a neologism such as inlier would be acceptable.
And interestingly enough, "inlier" may have been been in use for ~150 years. (see related entries)
There is a good English word of Latin origin for what is left over after something has been removed: residuum. It has the advantage over 'inliers' of not implying that what is left is more true than the data you had to start with - a fact always to be borne in mind when one is tempted to censor one's data.