Paper by Haoyue Dai*, Ignavier Ng*, Yujia Zheng, Zhengqing Gao, Kun Zhang. Appears at AISTATS 2024.
The first to tackle local causal discovery in cyclic models. By independent subspace analysis, all the local causal structures and coefficients in the equivalence class are identified (intersecting cycles are aslo allowed). A regression-based variant is given for acyclic cases.
Codes will be ready soon..
If you use this code for your research, please cite our paper:
@inproceedings{dai2024local,
title={Local Causal Discovery with Linear non-Gaussian Cyclic Models},
author={Haoyue Dai and Ignavier Ng and Yujia Zheng and Zhengqing Gao and Kun Zhang},
booktitle={International Conference on Artificial Intelligence and Statistics},
year={2024},
organization={PMLR}
}