We should consider renaming the package, since the focus is similarity instead of distance.
Although we should discuss whether we give the users the option to switch between sim & dist (most literature deals more with similarity tho)
In the current state we always calculate similarity. We even changed the min biparite matching from Jia et al. to max matching, to yield max similarity instead of min distance.
We should consider renaming the package, since the focus is similarity instead of distance. Although we should discuss whether we give the users the option to switch between sim & dist (most literature deals more with similarity tho) In the current state we always calculate similarity. We even changed the min biparite matching from Jia et al. to max matching, to yield max similarity instead of min distance.