AI Paper Atlas

Research map / Probabilistic and scientific modeling / Causal discovery

Structural Estimation of Partially Observed Linear Non-Gaussian Acyclic Model: A Practical Approach with Identifiability

Songyao Jin, Feng Xie, Guangyi Chen, Biwei Huang, Zhengming Chen, Xinshuai Dong, Kun Zhang

ICLR 2024 · Probabilistic and scientific modeling › Causal discovery · approach: Constraint-based structure learning with latents

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's abstract only. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Proposes a structural estimation method for partially observed linear non-Gaussian acyclic models (POLNAM) that recovers causal structure among measured and latent variables with identifiability guarantees.

Problem

Causal discovery over measured variables is fragile in the presence of latent variables, and existing latent-confounder methods often rely on strong structural assumptions.

Approach

Considers a general linear non-Gaussian acyclic model in which measured and latent variables jointly form the causal structure, and develops a practical estimation approach with identifiability results for this setting, without imposing strong assumptions on the latent structure.

Experimental setup

Builds on: LiNGAM

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