AI Paper Atlas

Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes

Calibrated Approximate Bayesian Inference

Hanwen Xing, Geoff Nicholls, Jeong Lee

ICML 2019 · Probabilistic and scientific modeling › Uncertainty and Gaussian processes · approach: Simulation-based inference · 4 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 AIS (Algorithm 2) instead of the importance samplercoverage estimation · Ising modeleffective sample size: 22 -> 180Improved
vary importance sampler threshold rho (ablation)Importance sampler (Lee et al.) · Ising modelcoverage estimateNo clear effect
use regression (BART, Algorithm 3) instead of AIScoverage estimation · Ising modelestimated coverage: 0.545 -> 0.565No clear effect
calibrate toroidal approximation versus exact credible set checktoroidal approximation · Ising modelcoverage of 95% credible set: 0.518Hurt
change summary statistics from T(y) (72-dim) to H(y) (17 order statistics) (ablation)Probit BART · Dirichlet process random effectsestimated coverage: 0.262 -> 0.308No clear effect
fit BART on quarter subsets of the training data (ablation)Probit BART M2 · Dirichlet process random effectsestimated coverage: 0.285, 0.322, 0.347, 0.270No clear effect
use BIC Laplace approximation to marginal likelihoodapproximate posterior over partitions · Dirichlet process random effectscoverage of 95% credible set: 0.262Hurt
use AIS (iterations to 60) instead of fewer iterationsAIS · Ising modelestimated coverage: 0.545Improved

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