Who can see your viewing activity?
Thanks for the talk. Why was the JL projection (and not other kinds) chosen here? Is there something special about JL for MST and FL?
I see — that makes sense.
Does your algorithm give an asymptotic speedup over just finding the MST exactly in the worst case doubling dimension (log n), or do you need it to be smaller?
Feel free to unmute yourself if you have any questions :)
asaf noar ribe program