AI Paper Atlas

Research map / Federated learning and privacy / Differential privacy

Differentially Private Cross-Silo Recommendation from Implicit Feedback

Xun Ran, Qingqing Ye, Xin Huang, Jianliang Xu, Haibo Hu

ICML 2026 · Federated learning and privacy › Differential privacy · approach: Differentially private federated learning

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add complementary loss and symmetric noise to strawman objective perturbation (ablation)DPIMF · ML-10MHR@10: DPIMFstr exhibits the worst performance due to high-variance noiseImproved
add symmetric noise (upper-triangular perturbation) to the complementary loss (ablation)DPIMFcom · ML-10M (alpha0=0.8)HR: 2.2% higher average HRImproved
set alpha0=1 (no second-order perturbation) (ablation)DPIMFsym · ML-10MHR@10: performs best across all epsImproved
add Poisson importance sampling (privacy amplification) to DPIMF (ablation)DPIMF · ML-10M, HR@10 vs epsHR@10: advantage more pronounced for eps < 5Improved
use privacy-aware importance sampling instead of uniform sampling (ablation)DPIMF-IS · ML-10M (2,000 users, d=20)sampling weights / variance of linear coefficient: weights closely match utility-optimal solution; uniform markedly differentImproved
replace DPLCF with DPIMF-ISprivate item profile on ML-10M · ML-10MHR: 18.3%Improved
replace LDPICF with DPIMF-ISprivate item profile on ML-10M · ML-10MHR: 8.3%Improved
replace DPLCF with DPIMF-ISprivate item profile on ML-10M · ML-10MNDCG: 12.1%Improved
replace LDPICF with DPIMF-ISprivate item profile on ML-10M · ML-10MNDCG: 5.4%Improved
replace SGDMF with DPIMF-IS (one-shot objective perturbation vs DP-SGD)DP recommender · ML-10M, eps=0.5HR@10: 0.15 -> 0.33Improved
replace EFVAE+DP-Adam with DPIMF-ISDP recommender · ML-10M, eps=0.5HR@10: 0.11 -> 0.33Improved
replace SGDMF with DPIMF-ISDP recommender · ML-10M, eps=10HR@10: 0.48 -> 0.54Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy