자료유형  

 학위논문

Control Number  

0017161449

International Standard Book Number  

9798383729649

Dewey Decimal Classification Number  

621.3

Main Entry-Personal Name  

Yao, Yuhang.

Publication, Distribution, etc. (Imprint  

[S.l.] : Carnegie Mellon University., 2024

Publication, Distribution, etc. (Imprint  

Ann Arbor : ProQuest Dissertations & Theses, 2024

Physical Description  

173 p.

General Note  

Source: Dissertations Abstracts International, Volume: 86-02, Section: B.

General Note  

Advisor: Joe-Wong, Carlee.

Dissertation Note  

Thesis (Ph.D.)--Carnegie Mellon University, 2024.

Summary, Etc.  

요약Graph neural networks aim to learn representations of graph-structured data that capture features associated with graph nodes as well as edges between the nodes. The Graph data, however, can be too large to be trained on a single server or naturally exist on multiple local clients. Strict data protection regulations such as the General Data Protection Regulation (GDPR) in Europe and Payment Aggregators and Payment Gateways (PAPG) in India also require that private data only be stored in local clients. Users may also not want to share their personal data with the server. Federated learning shows promise for preserving user privacy while training accurate models training models on data stored at multiple clients. Federated graph learning is then becoming an emerging topic with practical challenges. However, compared with federated learning with local data heterogeneity, there is no clear identification of key challenges in federated graph learning. Derived from real-world applications, we point out three unique challenges for federated graph learning: cross-client edges, cross-device heterogeneity with limited local data and temporal heterogeneity across clients. We then respectively introduce FedGCN, FedRule, and FedLink to overcome the challenges. We finally provide the vision for the future direction of such a topic.

Subject Added Entry-Topical Term  

Computer engineering.

Subject Added Entry-Topical Term  

Statistics.

Subject Added Entry-Topical Term  

Information technology.

Index Term-Uncontrolled  

Federated graph learning

Index Term-Uncontrolled  

Graph neural networks

Index Term-Uncontrolled  

Private data

Index Term-Uncontrolled  

Graph nodes

Added Entry-Corporate Name  

Carnegie Mellon University Electrical and Computer Engineering

Host Item Entry  

Dissertations Abstracts International. 86-02B.

Electronic Location and Access  

로그인을 한후 보실 수 있는 자료입니다.

Control Number  

joongbu:658293