A polynomial-time differentially private algorithm approximates every cut in a graph within (1+γ) multiplicative and n^{1.25+o(1)} additive error, breaking the previous n^{1.5} barrier.
Bencz\' u r and David R
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.DS 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Breaking the $n^{1.5}$ Additive Error Barrier for Private and Efficient Graph Sparsification via Private Expander Decomposition
A polynomial-time differentially private algorithm approximates every cut in a graph within (1+γ) multiplicative and n^{1.25+o(1)} additive error, breaking the previous n^{1.5} barrier.