We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

stat.ME

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Statistics > Methodology

Title: Ancestor regression in linear structural equation models

Abstract: We present a new method for causal discovery in linear structural equation models. We propose a simple "trick" based on statistical testing in linear models that shall distinguish between ancestors and non-ancestors of any given variable. Naturally, this can then be extended to estimating full graphs. Unlike many methods, we provide explicit error control for false causal discovery, at least asymptotically. This holds true even under Gaussianity where various methods fail due to non-identifiable structures.
Subjects: Methodology (stat.ME); Statistics Theory (math.ST)
Cite as: arXiv:2205.08925 [stat.ME]
  (or arXiv:2205.08925v1 [stat.ME] for this version)

Submission history

From: Christoph Schultheiss [view email]
[v1] Wed, 18 May 2022 13:45:37 GMT (312kb,D)

Link back to: arXiv, form interface, contact.