AI Paper Atlas

Research map / Probabilistic and scientific modeling / Variational and generative sampling

Pareto Smoothed Importance Sampling

Aki Vehtari, Daniel Simpson, Andrew Gelman, Yuling Yao, Jonah Gabry

JMLR 2024 · Probabilistic and scientific modeling › Variational and generative sampling · approach: Importance sampling and quadrature

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
replace ordinary IS with PSISimportance sampling · Exponential target, Exponential(theta) proposalsRMSE ratio: PSIS always better than ISImproved
replace truncated IS with PSISimportance sampling · Exponential(2) proposal (k-hat about 0.5)RMSE: TIS slightly better at intermediate sample sizesHurt
add PSIS correction to split-normal approximationSplit-normal · Logistic GP density estimation (Galaxy)KL divergence to MCMC: small KL-divergence already when S is 100Improved
use plain IS instead of PSISSplit-normal · Logistic GP density estimation (Galaxy)KL variability: mean k-hat 0.43 (S=100) and 0.55 (S=10^4)Hurt
add split scaling to the normal approximationNormal approximation · Logistic GP density estimation (Galaxy)mean Pareto k-hat: 0.60 -> 0.55Improved
use PSIS-LOO instead of exact LOOLeave-one-out cross-validation · 105 protein expression modelscomputation time: 125 days -> less than one minuteImproved
switch from Gaussian to Student-t observation modelLinear model for protein expression · 105 protein expression modelsnumber of cases with k-hat > 0.7: 352 -> 53Improved
add exact LOO for cases with k-hat > 0.7PSIS-LOO · 105 protein expression modelsLOO accuracy: 42 hours extra computationImproved
increase dimension D from 1 to 1024Importance sampling with Student-t proposal · D-dimensional normal target, S=10000effective sample size: D from 1 to 1024Hurt
apply sample-size dependent k-hat thresholdPSIS diagnostic · k-hat thresholdthreshold: S=100 -> 0.5Mixed

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