Reproduce “Sort, Partition, Randomize: Optimal Binary Hypothesis Testing under Local Differential Privacy” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Elena Ghazi, Jawad Nasser, Flavio Calmon, Ibrahim Issa
The authors show that finite-alphabet binary testing under pure non-interactive ε-LDP has an optimal mechanism formed by sorting symbols by likelihood ratio, grouping adjacent symbols, and applying randomized response to group labels. A dynamic program optimizes f-divergence objectives in O(ℓk²) time. For k=6, its KL utility matches the KOV optimum within 6.1×10^-14; this is an empirical check, but the theorem establishes optimality.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.