Paper Image

Causal Structure Learning via Matrix Factorization

Published on:

1 November 2023

Primary Category:

Machine Learning

Paper Authors:

Yunfeng Cai,

Xu Li,

Minging Sun,

Ping Li

Bullets

Key Details

Proposes Cholesky factorization based method (CDCF) to learn causal DAGs

CDCF runs in O(p3) time, faster than prior methods

CDCF has guarantees for exact recovery under assumptions

CDCF+ extends the method to handle latent variables

Experiments show state-of-the-art performance on synthetic and real data

AI generated summary

Causal Structure Learning via Matrix Factorization

This paper proposes a fast algorithm to learn the causal structure from observational data, by factorizing the covariance matrix. It runs in cubic time and has guarantees for exact recovery. The method is extended to handle latent variables.

Answers from this paper

Comments

No comments yet, be the first to start the conversation...

Sign up to comment on this paper

Sign Up