arXiv:2511. 14441v2 Announce Type: replace-cross Abstract: To distinguish Markov equivalent graphs in causal discovery, it is necessary to restrict the structural causal model.
Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv Machine Learning.
arXiv:2606. 08196v1 Announce Type: cross Abstract: We study causal discovery from observational data when some variables are hidden and the data-generating process follows a location-scale noise model (LSNM).
arXiv:2411.05625v2 Announce Type: replace Abstract: We propose a new approach to falsify causal discovery algorithms without ground truth, which is based on testing the causal model on a variable pai...
arXiv:2607. 11508v1 Announce Type: cross Abstract: Causal discovery, the process of recovering underlying causal structures from observational data, is a fundamental pursuit across scientific disciplines.
The paper investigates federated learning for linear non‑Gaussian acyclic models (LiNGAM), proposing the FedRCD family of algorithms that use higher‑order cumulants to enable privacy‑preserving causal discovery across distributed clients. It addresses limitations of existing federated methods, such as FedISHC’s failure under near‑symmetric noise, and introduces variants that balance communication rounds with algebraic noise handling. Experiments reveal that cumulant‑based federated approaches rank variables by a variance ladder induced by the DAG rather than by population asymmetry, and that marginal standardisation degrades performance while scale‑invariant DirectLiNGAM remains robust.
The paper introduces FedRCD, a family of federated causal discovery algorithms for linear non‑Gaussian acyclic models (LiNGAM). FedRCD overcomes limitations of the existing FedISHC method, which fails under near‑symmetric noise, by leveraging higher‑order cumulant tensors that aggregate across independent client data in a single communication round. Three variants of FedRCD trade off communication rounds against algebraic noise, with two exact federated counterparts of centralised HC and HC‑LiNGAM, and a single‑round variant that supports exact unlearning at any granularity.
arXiv:2601.01368v2 Announce Type: replace Abstract: Score-based causal discovery in the presence of unobserved confounders requires both a consistent scoring criterion and an efficient search over gr...