Current studentsExisting staffContact us
 

FedCSL: a scalable and accurate approach to federated causal structure learning

Simple item page
dc.contributor.authorGuo, X.
dc.contributor.authorYu, K.
dc.contributor.authorLiu, L.
dc.contributor.authorLi, J.
dc.contributor.conferenceThirty-Eighth AAAI Conference on Artificial Intelligence (20 Feb 2024 - 27 Feb 2024 : Vancouver)
dc.contributor.editorWooldridge, M.
dc.contributor.editorDy, J.
dc.contributor.editorNatarajan, S.
dc.date.issued2024
dc.description.abstractAs an emerging research direction, federated causal structure learning (CSL) aims at learning causal relationships from decentralized data across multiple clients while preserving data privacy. Existing federated CSL algorithms suffer from scalability and accuracy issues, since they require computationally expensive CSL algorithms to be executed at each client. Furthermore, in real-world scenarios, the number of samples held by each client varies significantly, and existing methods still assign equal weights to the learned structural information from each client, which severely harms the learning accuracy of those methods. To address these two limitations, we propose FedCSL, a scalable and accurate method for federated CSL. Specifically, FedCSL consists of two novel strategies: (1) a federated local-to-global learning strategy that enables FedCSL to scale to high-dimensional data for tackling the scalability issue, and (2) a novel weighted aggregation strategy that does not rely on any complex encryption techniques while preserving data privacy for tackling the accuracy issue. Extensive experiments on benchmark datasets, high-dimensional synthetic datasets and a real-world dataset verify the efficacy of the proposed FedCSL method. The source code is available at https://github.com/Xianjie-Guo/FedCSL.
dc.identifier.citationProceedings of the ... AAAI Conference on Artificial Intelligence. AAAI Conference on Artificial Intelligence, 2024 / Wooldridge, M., Dy, J., Natarajan, S. (ed./s), vol.38, iss.11, pp.12235-12243
dc.identifier.doi10.1609/aaai.v38i11.29113
dc.identifier.isbn9781577358879
dc.identifier.issn2159-5399
dc.identifier.issn2374-3468
dc.identifier.urihttps://hdl.handle.net/11541.2/38336
dc.language.isoen
dc.publisherAAAI Press
dc.publisher.placeUS
dc.relation.ispartofseries38, 2159-5399
dc.rightsCopyright 2024 Association for the Advancement of Artificial Intelligence Access Condition Notes: Accepted manuscript available after 1 April 2025
dc.source.urihttps://doi.org/10.1609/aaai.v38i11.29113
dc.subjectML: Causal Learning
dc.subjectML: Distributed Machine Learning & Federated Learning
dc.subjectML: Dimensionality Reduction/Feature Selection
dc.subjectDMKM: Scalability
dc.subjectParallel & Distributed Systems
dc.titleFedCSL: a scalable and accurate approach to federated causal structure learning
dc.typeConference paper
pubs.publication-statusPublished
ror.fileinfo12284557280001831 13284557270001831 Open Access Postprint
ror.mmsid9916847128101831

Files

Original bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
9916847128101831_12284557280001831_AM FedCSL.pdf
Size:
427.02 KB
Format:
Adobe Portable Document Format
Description:
Published version
Download

Collections