Differentially private top-k ranking from pairwise comparisons is minimax optimal, with exact rates sqrt(log n/(np)) + log n/(npε) under edge DP and sqrt(n log n/m) + n log n/(mε) under individual DP.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
math.ST 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Optimal Differentially Private Ranking from Pairwise Comparisons
Differentially private top-k ranking from pairwise comparisons is minimax optimal, with exact rates sqrt(log n/(np)) + log n/(npε) under edge DP and sqrt(n log n/m) + n log n/(mε) under individual DP.