1 paper
Henning Sulzbach, Ralph Neininger, Michael Drmota
We consider versions of the FIND algorithm where the pivot element used is the median of a subset chosen uniformly at random from the data. For the median selection we assume that…