Differentially Private and Federated Structure Learning in Bayesian Networks

Ghita Fassy El Fehri, Aurélien Bellet, Philippe Bastien
Proceedings of The 29th International Conference on Artificial Intelligence and Statistics, PMLR 300:4879-4887, 2026.

Abstract

Learning the structure of a Bayesian network from decentralized data poses two major challenges: (i) ensuring rigorous privacy guarantees for participants, and (ii) avoiding communication costs that scale poorly with dimensionality. In this work, we introduce Fed-Sparse-BNSL, a novel federated method for learning linear Gaussian Bayesian network structures that addresses both challenges. By combining differential privacy with greedy updates that target only a few relevant edges per participant, Fed-Sparse-BNSL efficiently uses the privacy budget while keeping communication costs low. Our careful algorithmic design preserves model identifiability and enables accurate structure estimation. Experiments on synthetic and real datasets demonstrate that Fed-Sparse-BNSL achieves utility close to non-private baselines while offering substantially stronger privacy and communication efficiency.

Cite this Paper

BibTeX
@InProceedings{pmlr-v300-el-fehri26a, title = { Differentially Private and Federated Structure Learning in Bayesian Networks }, author = {El Fehri, Ghita Fassy and Bellet, Aur{\'e}lien and Bastien, Philippe}, booktitle = {Proceedings of The 29th International Conference on Artificial Intelligence and Statistics}, pages = {4879--4887}, year = {2026}, editor = {Khan, Emtiyaz and Li, Yingzhen and Solin, Arno and Ramdas, Aaditya}, volume = {300}, series = {Proceedings of Machine Learning Research}, month = {02--05 May}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v300/main/assets/el-fehri26a/el-fehri26a.pdf}, url = {https://proceedings.mlr.press/v300/el-fehri26a.html}, abstract = { Learning the structure of a Bayesian network from decentralized data poses two major challenges: (i) ensuring rigorous privacy guarantees for participants, and (ii) avoiding communication costs that scale poorly with dimensionality. In this work, we introduce Fed-Sparse-BNSL, a novel federated method for learning linear Gaussian Bayesian network structures that addresses both challenges. By combining differential privacy with greedy updates that target only a few relevant edges per participant, Fed-Sparse-BNSL efficiently uses the privacy budget while keeping communication costs low. Our careful algorithmic design preserves model identifiability and enables accurate structure estimation. Experiments on synthetic and real datasets demonstrate that Fed-Sparse-BNSL achieves utility close to non-private baselines while offering substantially stronger privacy and communication efficiency. } }
Endnote
%0 Conference Paper %T Differentially Private and Federated Structure Learning in Bayesian Networks %A Ghita Fassy El Fehri %A Aurélien Bellet %A Philippe Bastien %B Proceedings of The 29th International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2026 %E Emtiyaz Khan %E Yingzhen Li %E Arno Solin %E Aaditya Ramdas %F pmlr-v300-el-fehri26a %I PMLR %P 4879--4887 %U https://proceedings.mlr.press/v300/el-fehri26a.html %V 300 %X Learning the structure of a Bayesian network from decentralized data poses two major challenges: (i) ensuring rigorous privacy guarantees for participants, and (ii) avoiding communication costs that scale poorly with dimensionality. In this work, we introduce Fed-Sparse-BNSL, a novel federated method for learning linear Gaussian Bayesian network structures that addresses both challenges. By combining differential privacy with greedy updates that target only a few relevant edges per participant, Fed-Sparse-BNSL efficiently uses the privacy budget while keeping communication costs low. Our careful algorithmic design preserves model identifiability and enables accurate structure estimation. Experiments on synthetic and real datasets demonstrate that Fed-Sparse-BNSL achieves utility close to non-private baselines while offering substantially stronger privacy and communication efficiency.
APA
El Fehri, G.F., Bellet, A. & Bastien, P.. (2026). Differentially Private and Federated Structure Learning in Bayesian Networks . Proceedings of The 29th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 300:4879-4887 Available from https://proceedings.mlr.press/v300/el-fehri26a.html.

Related Material