AI Paper Atlas

Research map / Federated learning and privacy / Differential privacy

Particle Filter for Bayesian Inference on Privatized Data

Yu-Wei Chen, Pranav Sanghi, Jordan Awan

JMLR 2026 · Federated learning and privacy › Differential privacy · approach: Local-DP and private estimation · 2 citations (OpenAlex)

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
use DP-PF instead of DP-DA-MCMCprivate posterior sampler · Linear regression non-conjugate prior, eps=0.5, n=500seconds per ESS: 1173.600 -> 0.638Improved
use DP-PF instead of DP-DA-MCMCprivate posterior sampler · Linear regression non-conjugate prior, eps=0.5, n=50090% HPD width of beta_1: 3.009 -> 2.097 (Reject-ABC 2.186)Improved
use DP-PF instead of DP-Reject-ABCprivate posterior sampler · Linear regression non-conjugate prior, eps=1.0, n=500seconds per ESS: 70.577 -> 1.696Improved
use DP-PF instead of DP-Reject-ABC at large n and epsprivate posterior sampler · Location-scale normal, eps=2.0, n=1000seconds per ESS: 5.74 -> 0.07Improved
use DP-PF instead of DP-DA-MCMC with conjugate priorprivate posterior sampler · Linear regression conjugate prior, eps=0.5, n=500seconds per ESS: 45.503 -> 0.868Improved
use DP-PF instead of DP-Reject-ABC with conjugate prior at eps=2, n=300private posterior sampler · Linear regression conjugate prior, eps=2, n=300seconds per ESS: 17.053 -> 0.591Improved
use DP-PF instead of DP-DA-MCMC with M=2 histogram layersprivate posterior sampler · Gaussian mixture model, M=2posterior recovery of parametersImproved
exponential instead of linear epsilon_t schedule (ablation)DP-PF · Linear regression non-conjugate prior, eps=2.0, n=500seconds per ESS: 7.481 -> 14.021Hurt
exponential instead of linear epsilon_t schedule (ablation)DP-PF · Linear regression non-conjugate prior, eps=0.5, n=300posterior mean of beta_1: 1.973 -> 1.970No clear effect
exponential instead of linear epsilon_t schedule (ablation)DP-PF · Linear regression conjugate prior, eps=2.0, n=500seconds per ESS: 11.657 -> 22.512Hurt
increase number of particles N beyond 1000 (ablation)DP-PF · Linear regression CI coverage, eps>=0.595% CI coverage and width: coverage around 0.95 and CI width below 0.05 for N > 1000Improved
smaller privacy budget eps (ablation)DP-PF · Linear regression CI coverageparticles needed for coverageHurt

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