ParceLiNGAM: a causal ordering method robust against latent confounders
From MaRDI portal
Publication:5378311
DOI10.1162/NECO_A_00533zbMATH Open1410.68325arXiv1303.7410OpenAlexW2135404705WikidataQ48952309 ScholiaQ48952309MaRDI QIDQ5378311FDOQ5378311
Authors: Tatsuya Tashiro, Shohei Shimizu, Aapo Hyvärinen, Takashi Washio 
Publication date: 12 June 2019
Published in: Neural Computation (Search for Journal in Brave)
Abstract: We consider learning a causal ordering of variables in a linear non-Gaussian acyclic model called LiNGAM. Several existing methods have been shown to consistently estimate a causal ordering assuming that all the model assumptions are correct. But, the estimation results could be distorted if some assumptions actually are violated. In this paper, we propose a new algorithm for learning causal orders that is robust against one typical violation of the model assumptions: latent confounders. The key idea is to detect latent confounders by testing independence between estimated external influences and find subsets (parcels) that include variables that are not affected by latent confounders. We demonstrate the effectiveness of our method using artificial data and simulated brain imaging data.
Full work available at URL: https://arxiv.org/abs/1303.7410
Recommendations
- Causality in linear nongaussian acyclic models in the presence of latent Gaussian confounders
- DirectLiNGAM: a direct method for learning a linear non-Gaussian structural equation model
- Finding a causal ordering via independent component analysis
- Pairwise likelihood ratios for estimation of non-Gaussian structural equation models
- A linear non-Gaussian acyclic model for causal discovery
Cites Work
- Causation, prediction, and search
- Title not available (Why is that?)
- 10.1162/153244303768966085
- Title not available (Why is that?)
- Causal diagrams for empirical research
- A linear non-Gaussian acyclic model for causal discovery
- Analyse générale des liaisons stochastiques: etude particulière de l'analyse factorielle linéaire
- Title not available (Why is that?)
- DirectLiNGAM: a direct method for learning a linear non-Gaussian structural equation model
- Pairwise likelihood ratios for estimation of non-Gaussian structural equation models
- Estimation of causal effects using linear non-Gaussian causal models with hidden variables
- Causality in linear nongaussian acyclic models in the presence of latent Gaussian confounders
Cited In (12)
- Causal discovery in heavy-tailed models
- Identifiability of latent-variable and structural-equation models: from linear to nonlinear
- An integrated causal path identification method
- Pairwise likelihood ratios for estimation of non-Gaussian structural equation models
- Finding a causal ordering via independent component analysis
- Estimating bounds on causal effects in high-dimensional and possibly confounded systems
- Multi-Trek Separation in Linear Structural Equation Models
- Learning linear non-Gaussian causal models in the presence of latent variables
- Causality in linear nongaussian acyclic models in the presence of latent Gaussian confounders
- CAM: causal additive models, high-dimensional order search and penalized regression
- Learning instrumental variables with structural and non-Gaussianity assumptions
- A causal discovery algorithm based on the prior selection of leaf nodes
Uses Software
This page was built for publication: ParceLiNGAM: a causal ordering method robust against latent confounders
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5378311)
