Comparing Population Means Under Local Differential Privacy: With Significance and Power

Authors: Bolin Ding, Harsha Nori, Paul Li, Joshua Allen

AAAI 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We report experimental results on real-world datasets to verify the effectiveness of our approaches.
Researcher Affiliation Industry Bolin Ding, Harsha Nori, Paul Li, Joshua Allen {bolind, hanori, paul.li, joshuaa}@microsoft.com Microsoft, One Microsoft Way, Redmond, WA 98052
Pseudocode Yes Algorithm 1: Tbin ε : Transformation-based LDP Test; Algorithm 2: Tmix ε : For Hybrid Privacy Requirements
Open Source Code No The paper does not contain any statements about releasing open-source code or links to a code repository for the described methodology.
Open Datasets No There are 20 million users in this real-world dataset. Each user has a counter with the value in [0, 15000], i.e., m = 15000... We draw samples from control/treatment with equal sizes n A = n B. The paper mentions a 'real-world dataset' but does not provide a name, specific link, or citation for public access.
Dataset Splits No The paper discusses 'samples' and 'sample sizes' (n A, n B) but does not specify dataset splits for training, validation, or testing in the typical machine learning sense (e.g., percentages or counts for each split).
Hardware Specification No The paper does not provide any specific details about the hardware (e.g., GPU/CPU models, memory, cloud instances) used to run the experiments.
Software Dependencies No The paper does not specify any software dependencies with version numbers.
Experiment Setup Yes Parameters: privacy budget ε and significance level α... We vary the privacy parameter ε from 0.5 to 5. The pre-specified significance level α = 0.05, and the null hypothesis is H0 : μA μB = 0.