Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes
Expectation Propagation as a Way of Life: A Framework for Bayesian Inference on Partitioned Data
Aki Vehtari, Andrew Gelman, Tuomas Sivula, Pasi Jylänki, Dustin Tran, Swupnil Sahai, Paul Blomstedt, John P. Cunningham, David Schiminovich, Christian P. Robert
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 changed | Where | Measured | Effect |
|---|---|---|---|
| increase number of EP sites from 1 to 30 with parallel updates | Distributed EP on astronomy mixture model · Astronomy FUV vs i100 | computation time: 96% decrease in computation time | Improved |
| increase number of partitions K from 2 to 64 | Distributed EP · Simulated hierarchical logistic regression | MSE and KL of approximation: K=2 vs K=64, final accuracy declines as K grows | Hurt |
| use EP with cavity priors instead of consensus Monte Carlo with fractioned prior | Distributed inference · Simulated hierarchical logistic regression | MSE and KL divergence: EP better than consensus MC for all K = 2,4,8,16,32,64 | Improved |
| increase number of sites in astronomy fits | Distributed EP · Astronomy FUV vs i100 | local fit quality: K=30 differs at 82 degrees, K=10 differs at 194 degrees | Mixed |
Closest papers by meaning
- Fast Expectation Propagation for Heteroscedastic, Lasso-Penalized, and Quantile Regression (JMLR 2023 · similarity 0.90)
- Distributed Bayesian Learning with Stochastic Natural Gradient Expectation Propagation and the Posterior Server (JMLR 2017 · similarity 0.89)
- Fearless Stochasticity in Expectation Propagation (NeurIPS 2024 · similarity 0.89)
- Automated Scalable Bayesian Inference via Hilbert Coresets (JMLR 2019 · similarity 0.87)
- Fast Bayesian Coresets via Subsampling and Quasi-Newton Refinement (NeurIPS 2022 · similarity 0.87)
- Bayes-Newton Methods for Approximate Bayesian Inference with PSD Guarantees (JMLR 2023 · similarity 0.87)
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.
