서브메뉴
검색
Advancing Graph Neural Networks for Complex Data: A Perspective Beyond Homophily
Advancing Graph Neural Networks for Complex Data: A Perspective Beyond Homophily
상세정보
- 자료유형
- 학위논문 서양
- 최종처리일시
- 20250211153018
- ISBN
- 9798384046165
- DDC
- 004
- 저자명
- Zhu, Jiong.
- 서명/저자
- Advancing Graph Neural Networks for Complex Data: A Perspective Beyond Homophily
- 발행사항
- [Sl] : University of Michigan, 2024
- 발행사항
- Ann Arbor : ProQuest Dissertations & Theses, 2024
- 형태사항
- 257 p
- 주기사항
- Source: Dissertations Abstracts International, Volume: 86-03, Section: B.
- 주기사항
- Advisor: Koutra, Danai.
- 학위논문주기
- Thesis (Ph.D.)--University of Michigan, 2024.
- 초록/해제
- 요약Graph Neural Networks (GNNs) have demonstrated significant potential in extending the empirical success of deep learning from Euclidean spaces to non-Euclidean, graph-structured data. These models operate on versatile relational networks to extract meaningful representations, enabling a wide range of downstream applications such as friend recommendations, fraud detection, and bioinformatics. A key principle underlying many real-world networks-and often implicitly leveraged by GNN models-is homophily, whereby linked nodes often belong to the same class or have similar features ("birds of a feather flock together"). However, real-world settings where "opposites attract" also exist, resulting in networks characterized by heterophily, where linked nodes are likely from different classes or possess dissimilar features. How do GNNs perform in settings where the homophily is weak? Is the reduced accuracy of GNNs under heterophily related to their limitations in other performance aspects such as robustness, fairness, and scalability? This dissertation aims to push forward the state-of-the-art in GNNs by addressing these questions. In Part I, I focus on understanding and improving GNN accuracy beyond homophily. I first examine the limitations of existing GNN models under heterophily for semi-supervised node classification tasks, introducing key design strategies and new methods that significantly enhance learning from the graph structure on such datasets. Furthermore, I extend this analysis to link prediction tasks by formalizing the definitions of non-homophilic link prediction based on feature similarity, analyzing how different link prediction encoders and decoders adapt to varying levels of feature similarity, and introducing designs for improved performance. In Part II, I explore the implications of heterophily on other critical GNN research objectives beyond accuracy, including adversarial robustness, algorithmic fairness, and distributed scalability. My findings reveal that addressing heterophily not only enhances GNN accuracy but also improves their robustness and fairness, making them more suitable for deployment in complex real-world applications. Additionally, I show that heterophily is key to streamlining distributed training of GNNs on massive graphs, reducing communication overhead and improving efficiency and scalability. Overall, this dissertation advances the field of GNNs by moving beyond homophily, offering new insights and methodologies for handling heterophilous graph datasets, and demonstrating the broader benefits of these advancements in terms of robustness, fairness, and scalability.
- 일반주제명
- Computer science
- 일반주제명
- Computer engineering
- 일반주제명
- Information technology
- 키워드
- Heterophily
- 키워드
- Machine learning
- 키워드
- Data mining
- 키워드
- Homophily
- 기타저자
- University of Michigan Computer Science & Engineering
- 기본자료저록
- Dissertations Abstracts International. 86-03B.
- 전자적 위치 및 접속
- 로그인 후 원문을 볼 수 있습니다.
MARC
008250123s2024 us c eng d■001000017164575
■00520250211153018
■006m o d
■007cr#unu||||||||
■020 ▼a9798384046165
■035 ▼a(MiAaPQ)AAI31631558
■035 ▼a(MiAaPQ)umichrackham005704
■040 ▼aMiAaPQ▼cMiAaPQ
■0820 ▼a004
■1001 ▼aZhu, Jiong.
■24510▼aAdvancing Graph Neural Networks for Complex Data: A Perspective Beyond Homophily
■260 ▼a[Sl]▼bUniversity of Michigan▼c2024
■260 1▼aAnn Arbor▼bProQuest Dissertations & Theses▼c2024
■300 ▼a257 p
■500 ▼aSource: Dissertations Abstracts International, Volume: 86-03, Section: B.
■500 ▼aAdvisor: Koutra, Danai.
■5021 ▼aThesis (Ph.D.)--University of Michigan, 2024.
■520 ▼aGraph Neural Networks (GNNs) have demonstrated significant potential in extending the empirical success of deep learning from Euclidean spaces to non-Euclidean, graph-structured data. These models operate on versatile relational networks to extract meaningful representations, enabling a wide range of downstream applications such as friend recommendations, fraud detection, and bioinformatics. A key principle underlying many real-world networks-and often implicitly leveraged by GNN models-is homophily, whereby linked nodes often belong to the same class or have similar features ("birds of a feather flock together"). However, real-world settings where "opposites attract" also exist, resulting in networks characterized by heterophily, where linked nodes are likely from different classes or possess dissimilar features. How do GNNs perform in settings where the homophily is weak? Is the reduced accuracy of GNNs under heterophily related to their limitations in other performance aspects such as robustness, fairness, and scalability? This dissertation aims to push forward the state-of-the-art in GNNs by addressing these questions. In Part I, I focus on understanding and improving GNN accuracy beyond homophily. I first examine the limitations of existing GNN models under heterophily for semi-supervised node classification tasks, introducing key design strategies and new methods that significantly enhance learning from the graph structure on such datasets. Furthermore, I extend this analysis to link prediction tasks by formalizing the definitions of non-homophilic link prediction based on feature similarity, analyzing how different link prediction encoders and decoders adapt to varying levels of feature similarity, and introducing designs for improved performance. In Part II, I explore the implications of heterophily on other critical GNN research objectives beyond accuracy, including adversarial robustness, algorithmic fairness, and distributed scalability. My findings reveal that addressing heterophily not only enhances GNN accuracy but also improves their robustness and fairness, making them more suitable for deployment in complex real-world applications. Additionally, I show that heterophily is key to streamlining distributed training of GNNs on massive graphs, reducing communication overhead and improving efficiency and scalability. Overall, this dissertation advances the field of GNNs by moving beyond homophily, offering new insights and methodologies for handling heterophilous graph datasets, and demonstrating the broader benefits of these advancements in terms of robustness, fairness, and scalability.
■590 ▼aSchool code: 0127.
■650 4▼aComputer science
■650 4▼aComputer engineering
■650 4▼aInformation technology
■653 ▼aGraph Neural Networks
■653 ▼aHeterophily
■653 ▼aMachine learning
■653 ▼aData mining
■653 ▼aHomophily
■690 ▼a0984
■690 ▼a0800
■690 ▼a0489
■690 ▼a0464
■71020▼aUniversity of Michigan▼bComputer Science & Engineering.
■7730 ▼tDissertations Abstracts International▼g86-03B.
■790 ▼a0127
■791 ▼aPh.D.
■792 ▼a2024
■793 ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17164575▼nKERIS▼z이 자료의 원문은 한국교육학술정보원에서 제공합니다.


