Research map / Probabilistic and scientific modeling / Bayesian optimization

Gaussian Ensemble Belief Propagation for Efficient Inference in High-Dimensional, Black-box Systems

ICLR 2025 · Probabilistic and scientific modeling › Bayesian optimization · approach: High-dimensional Bayesian optimization

Read the paperCheck your idea against 108,967 papers

AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Introduces Gaussian Ensemble Belief Propagation (GEnBP), combining the Ensemble Kalman Filter with Gaussian Belief Propagation to update prior ensembles into posterior samples via low-rank local messages on a graph.

Problem

Efficient inference in high-dimensional, black-box models where posterior computation is costly and the model is accessed only as a function evaluator.

Approach

Combines EnKF's ensemble-based sample updates with GaBP's local message passing over a graph, so that ensembles are updated using low-rank local messages on edges instead of a full joint Gaussian update.

Experimental setup

Builds on: Ensemble Kalman Filter Gaussian Belief Propagation

Closest papers by meaning

• Posterior Inference with Diffusion Models for High-dimensional Black-box Optimization (ICML 2025 · similarity 0.83)
• End-to-end Learning of Gaussian Mixture Priors for Diffusion Sampler (ICLR 2025 · similarity 0.83)
• Computation-Aware Gaussian Processes: Model Selection And Linear-Time Inference (NeurIPS 2024 · similarity 0.83)
• BGDB: Bernoulli-Gaussian Decision Block with Improved Denoising Diffusion Probabilistic Models (AAAI 2025 · similarity 0.83)
• Gated Inference Network: Inference and Learning State-Space Models (NeurIPS 2024 · similarity 0.83)
• Conditioning Gaussian Processes on Almost Anything (NeurIPS 2026 · similarity 0.83)