Differentially Private Chi-squared Test by Unit Circle Mechanism

Authors: Kazuya Kakizaki, Kazuto Fukuchi, Jun Sakuma

ICML 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental 7. Experiment In this section, we evaluate the significance and the power of the respective mechanisms, input perturbation (Gaboardi et al., 2016), output perturbation (Yu et al., 2014), and unit circle mechanism for single hypothesis testing.
Researcher Affiliation Academia 1Department of Computer Science, University of Tsukuba, 1-1-1 Tennohdai, Tsukuba, Ibaraki, Japan 2JST CREST 3RIKEN Center for Advanced Intelligence Project.
Pseudocode Yes Algorithm 1 Unit Circle Mechanism Require: Sample set S, sig level α, privacy budget ϵ
Open Source Code No No explicit statement or link regarding the provision of open-source code for the described methodology was found.
Open Datasets No No concrete access information (specific link, DOI, repository name, formal citation with authors/year, or reference to established benchmark datasets) for a publicly available or open dataset was found. The paper describes generating synthetic data for its experiments.
Dataset Splits No No specific dataset split information (exact percentages, sample counts, citations to predefined splits, or detailed splitting methodology) for training, validation, or test sets was provided.
Hardware Specification No No specific hardware details (exact GPU/CPU models, processor types with speeds, memory amounts, or detailed computer specifications) used for running its experiments were provided.
Software Dependencies No No specific ancillary software details, such as library or solver names with version numbers, were provided to replicate the experiment.
Experiment Setup Yes We set the privacy parameter as ϵ = 0.1 and significance level as α = 0.05. For Monte Carlo sampling, we set the number of sampling as 1000 for MCIndep Lap with the laplace distribution, and 10000 for the other methods. We set the privacy parameter as ϵ = {0.1, 0.5}, the significance levels as α = 0.05, and the stop parameters as s1 = 2, s2 = 10.