You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Currently SLHC (and HDBSCAN) use MST restarting capability in order to provide additional edges that will connect the original KNN graph. However, since it uses the existing 1-nn primitive which is used for k-means), all of the n^2 distances are computed everytime, which is very wasteful.
These algorithms should see some speedup by computing the distances only across components/colors.
The text was updated successfully, but these errors were encountered:
This issue has been labeled inactive-30d due to no recent activity in the past 30 days. Please close this issue if no further response or action is needed. Otherwise, please respond with a comment indicating any updates or changes to the original issue and/or confirm this issue still needs to be addressed. This issue will be labeled inactive-90d if there is no activity in the next 60 days.
Currently SLHC (and HDBSCAN) use MST restarting capability in order to provide additional edges that will connect the original KNN graph. However, since it uses the existing 1-nn primitive which is used for k-means), all of the n^2 distances are computed everytime, which is very wasteful.
These algorithms should see some speedup by computing the distances only across components/colors.
The text was updated successfully, but these errors were encountered: