AI Paper Atlas

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

Probabilistic Unrolling: Scalable, Inverse-Free Maximum Likelihood Estimation for Latent Gaussian Models

Alexander Lin, Bahareh Tolooshams, Yves Atchade, Demba E. Ba

ICML 2023 · Probabilistic and scientific modeling › Variational and generative sampling · approach: Latent variable marginalization and estimators · 1 citations (OpenAlex)

Read the paper

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 matrix-inversion gradient EM with probabilistic unrollingNoisy AR parameter recovery · D=1,000training time: 41±0 s -> 8±0 sImproved
Replace matrix-inversion gradient EM with probabilistic unrollingNoisy AR parameter recovery · D=30,000training time: 4139±49 s -> 87±1 sImproved
Replace matrix-inversion gradient EM with probabilistic unrollingNoisy AR parameter recovery · D=30,000r(φ, φ*): 0.5±0.2 % -> 0.4±0.2 %Improved
Replace matrix-inversion gradient EM with probabilistic unrollingBayesian compressed sensing · NIST digits averagetraining time: 1481±19 s -> 21±0 sImproved
Replace matrix-inversion gradient EM with probabilistic unrollingBayesian compressed sensing · NIST digits averageNRMSE: 4.8±1.0 % -> 4.7±1.4 %No clear effect
Replace matrix-inversion gradient EM with probabilistic unrollingFactor-analysis collaborative filtering · MovieLens ML-1Mtime per cycle: 54 min, 42 s -> 5 min, 50 sImproved
Replace matrix-inversion gradient EM with probabilistic unrollingFactor-analysis collaborative filtering · MovieLens ML-1MRMSE: 0.8433 -> 0.8436No clear effect
Replace matrix-inversion gradient EM with probabilistic unrollingFactor-analysis collaborative filtering · MovieLens ML-10Mtime per cycle: 78 min, 36 s -> 12 min, 8 sImproved
Replace matrix-inversion gradient EM with probabilistic unrollingFactor-analysis collaborative filtering · MovieLens ML-10MRMSE: 0.7809 -> 0.7796Improved
Use probabilistic unrolling to fit the factor-analysis modelFactor-analysis collaborative filtering · MovieLens ML-25Mfeasibility under memory limitImproved

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