The tendency of high-dimensional data to contain points (hubs) that frequently occur in k-nearest-neighbor lists of other points.
From hub + -ness.
Select a language to see translations