AI Paper Atlas

Research map / Probabilistic and scientific modeling / Causal discovery

Integrating Markov Blanket Discovery into Causal Representation Learning for Domain Generalization

Naiyu Yin, Hanjing Wang, Yue Yu, Tian Gao, Amit Dhurandhar, Qiang Ji

ECCV 2024 · Probabilistic and scientific modeling › Causal discovery · approach: Causal representation identifiability · 1 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
select the full latent Markov Blanket set instead of only parent variables (ablation)CMBRL · Colored MNIST OODaccuracy (%): 69.8 -> 74.1Improved
use CMBRL instead of ERM, which fits spurious color correlationsColored MNIST · Colored MNIST OODaccuracy (%): 10.3 -> 74.1Improved
use CMBRL instead of ERM on in-distribution dataColored MNIST · Colored MNIST in-distributionaccuracy (%): 85.7 -> 76.4Hurt
use CMBRL instead of the second-best baseline iCaRL on CelebACelebA OOD · CelebA OOD (avg over 7 settings)accuracy (%): 71.5 -> 76.2Improved
use CMBRL instead of iCaRL on VLCS and PACSdomain generalization · VLCS and PACS (ResNet50)average accuracy (%): VLCS 81.8 -> 82.3; PACS 88.7 -> 89.1Improved
use NF-iVAE prior in phase I instead of VAE or iVAECMBRL · Colored MNISTMCC (identifiability)Improved
use CMBRL against IRM variants with only two training domainsColored MNIST · Colored MNIST OODaccuracy (%): 62.8 -> 74.1Improved

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