MLLGCOMEMar 13, 2019

Variational Bayesian Optimal Experimental Design

arXiv:1903.05480v3176 citations
Originality Incremental advance
AI Analysis

This work addresses a bottleneck in experimental design for researchers, offering incremental advances in efficiency and practicality.

The paper tackles the challenge of accurately estimating expected information gain in Bayesian optimal experimental design by introducing fast estimators based on amortized variational inference, achieving significant improvements in speed and accuracy over prior methods.

Bayesian optimal experimental design (BOED) is a principled framework for making efficient use of limited experimental resources. Unfortunately, its applicability is hampered by the difficulty of obtaining accurate estimates of the expected information gain (EIG) of an experiment. To address this, we introduce several classes of fast EIG estimators by building on ideas from amortized variational inference. We show theoretically and empirically that these estimators can provide significant gains in speed and accuracy over previous approaches. We further demonstrate the practicality of our approach on a number of end-to-end experiments.

Code Implementations1 repo
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes