Linear Causal Discovery with Interventional Constraints

Zhigao Guo, Feng Dong · 2025

Read Linear Causal Discovery with Interventional Constraints by Zhigao Guo, Feng Dong (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2510-26342.txt.

Description

Zhigao Guo and Feng Dong work on causal discovery, the task of learning which things cause which other things from data. They add a new tool called interventional constraints. These are plain rules about the sign of a cause, such as one protein turning another one on, rather than data from a direct test.

Guo and Dong show why older methods can get the sign wrong even when they find the right path between two factors. Their fix sets a limit on the total effect one factor may have on another, and they solve the task in two stages. Tests on real data sets gave models that matched known findings more closely, stayed easier to explain, and still turned up new links.

Catalog record

CREATORZhigao Guo, Feng Dong
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE94 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2510-26342

Browse the full library · About the Tower of Babel