Differentially Private Empirical Risk Minimization Revisited: Faster and More General

Authors: Di Wang, Minwei Ye, Jinhui Xu

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

Reproducibility Variable Result LLM Response
Research Type Experimental Due to space limit, we leave many details, proofs, and experimental studies in the supplement.
Researcher Affiliation Academia Di Wang Dept. of Computer Science and Engineering State University of New York at Buffalo Buffalo, NY 14260 dwang45@buffalo.edu Minwei Ye Dept. of Computer Science and Engineering State University of New York at Buffalo Buffalo, NY 14260 minweiye@buffalo.edu Jinhui Xu Dept. of Computer Science and Engineering State University of New York at Buffalo Buffalo, NY 14260 jinhui@buffalo.edu
Pseudocode Yes Algorithm 1 DP-SVRG(F r, x0, T, m, η, σ) Input: f(x, z) is G-Lipschitz and L-smooth. F r(x, D) is µ-strongly convex w.r.t ℓ2-norm. x0 is the initial point, η is the step size, T, m are the iteration numbers. 1: for s = 1, 2, , T do 2: x = xs 1 3: v = F( x) 4: xs 0 = x 5: for t = 1, 2, , m do 6: Pick is t [n] 7: vs t = f(xs t 1, zis t ) f( x, zis t ) + v + us t, where us t N(0, σ2Ip) 8: xs t = proxηr(xs t 1 ηvs t ) 9: end for 10: xs = 1 m Pm k=1 xs k 11: end for 12: return x T
Open Source Code No The paper does not provide any explicit statements about releasing source code or links to a code repository.
Open Datasets No The paper discusses 'dataset D = {z1, z2 , zn}' and 'large datasets' but does not provide specific access information (link, DOI, formal citation with authors/year) for any publicly available or open dataset.
Dataset Splits No The paper does not provide specific dataset split information (percentages, sample counts, citations to predefined splits, or detailed splitting methodology).
Hardware Specification No The paper does not provide specific hardware details (exact GPU/CPU models, processor types, or detailed computer specifications) used for running experiments.
Software Dependencies No The paper does not provide specific ancillary software details (e.g., library or solver names with version numbers).
Experiment Setup No The paper does not contain specific experimental setup details such as concrete hyperparameter values or training configurations.