Differentiable Structure Learning with Ancestral Constraints

Taiyu Ban, Changxin Rong, Xiangyu Wang, Lyuzhou Chen, Xin Wang, Derui Lyu, Qinrui Zhu, Huanhuan Chen
Proceedings of the 42nd International Conference on Machine Learning, PMLR 267:2801-2835, 2025.

Abstract

Differentiable structure learning of causal directed acyclic graphs (DAGs) is an emerging field in causal discovery, leveraging powerful neural learners. However, the incorporation of ancestral constraints, essential for representing abstract prior causal knowledge, remains an open research challenge. This paper addresses this gap by introducing a generalized framework for integrating ancestral constraints. Specifically, we identify two key issues: the non-equivalence of relaxed characterizations for representing path existence and order violations among paths during optimization. In response, we propose a binary-masked characterization method and an order-guided optimization strategy, tailored to address these challenges. We provide theoretical justification for the correctness of our approach, complemented by experimental evaluations on both synthetic and real-world datasets.

Cite this Paper

BibTeX
@InProceedings{pmlr-v267-ban25a, title = {Differentiable Structure Learning with Ancestral Constraints}, author = {Ban, Taiyu and Rong, Changxin and Wang, Xiangyu and Chen, Lyuzhou and Wang, Xin and Lyu, Derui and Zhu, Qinrui and Chen, Huanhuan}, booktitle = {Proceedings of the 42nd International Conference on Machine Learning}, pages = {2801--2835}, year = {2025}, editor = {Singh, Aarti and Fazel, Maryam and Hsu, Daniel and Lacoste-Julien, Simon and Berkenkamp, Felix and Maharaj, Tegan and Wagstaff, Kiri and Zhu, Jerry}, volume = {267}, series = {Proceedings of Machine Learning Research}, month = {13--19 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v267/main/assets/ban25a/ban25a.pdf}, url = {https://proceedings.mlr.press/v267/ban25a.html}, abstract = {Differentiable structure learning of causal directed acyclic graphs (DAGs) is an emerging field in causal discovery, leveraging powerful neural learners. However, the incorporation of ancestral constraints, essential for representing abstract prior causal knowledge, remains an open research challenge. This paper addresses this gap by introducing a generalized framework for integrating ancestral constraints. Specifically, we identify two key issues: the non-equivalence of relaxed characterizations for representing path existence and order violations among paths during optimization. In response, we propose a binary-masked characterization method and an order-guided optimization strategy, tailored to address these challenges. We provide theoretical justification for the correctness of our approach, complemented by experimental evaluations on both synthetic and real-world datasets.} }
Endnote
%0 Conference Paper %T Differentiable Structure Learning with Ancestral Constraints %A Taiyu Ban %A Changxin Rong %A Xiangyu Wang %A Lyuzhou Chen %A Xin Wang %A Derui Lyu %A Qinrui Zhu %A Huanhuan Chen %B Proceedings of the 42nd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2025 %E Aarti Singh %E Maryam Fazel %E Daniel Hsu %E Simon Lacoste-Julien %E Felix Berkenkamp %E Tegan Maharaj %E Kiri Wagstaff %E Jerry Zhu %F pmlr-v267-ban25a %I PMLR %P 2801--2835 %U https://proceedings.mlr.press/v267/ban25a.html %V 267 %X Differentiable structure learning of causal directed acyclic graphs (DAGs) is an emerging field in causal discovery, leveraging powerful neural learners. However, the incorporation of ancestral constraints, essential for representing abstract prior causal knowledge, remains an open research challenge. This paper addresses this gap by introducing a generalized framework for integrating ancestral constraints. Specifically, we identify two key issues: the non-equivalence of relaxed characterizations for representing path existence and order violations among paths during optimization. In response, we propose a binary-masked characterization method and an order-guided optimization strategy, tailored to address these challenges. We provide theoretical justification for the correctness of our approach, complemented by experimental evaluations on both synthetic and real-world datasets.
APA
Ban, T., Rong, C., Wang, X., Chen, L., Wang, X., Lyu, D., Zhu, Q. & Chen, H.. (2025). Differentiable Structure Learning with Ancestral Constraints. Proceedings of the 42nd International Conference on Machine Learning, in Proceedings of Machine Learning Research 267:2801-2835 Available from https://proceedings.mlr.press/v267/ban25a.html.

Related Material