서브메뉴
검색
Enhancing Networked Systems: A Comprehensive Approach to Robust and Privacy-Preserving Optimization Algorithms
Enhancing Networked Systems: A Comprehensive Approach to Robust and Privacy-Preserving Optimization Algorithms
상세정보
- 자료유형
- 학위논문 서양
- 최종처리일시
- 20250211151348
- ISBN
- 9798382843346
- DDC
- 621.3
- 저자명
- Ravi, Nikhil.
- 서명/저자
- Enhancing Networked Systems: A Comprehensive Approach to Robust and Privacy-Preserving Optimization Algorithms
- 발행사항
- [Sl] : Cornell University, 2024
- 발행사항
- Ann Arbor : ProQuest Dissertations & Theses, 2024
- 형태사항
- 205 p
- 주기사항
- Source: Dissertations Abstracts International, Volume: 85-12, Section: A.
- 주기사항
- Advisor: Scaglione, Anna.
- 학위논문주기
- Thesis (Ph.D.)--Cornell University, 2024.
- 초록/해제
- 요약This dissertation explores two pivotal aspects of networked systems operating in increasingly complex and adversarial environments: robust decentralized optimization and privacy preservation. Part I delves into the development of decentralized optimization algorithms, addressing the critical challenge of ensuring system robustness against adversarial threats. It investigates strategies to secure the integrity of consensus and optimization processes in distributed networks, where each node or agent contributes to a global objective without centralized coordination. This exploration underscores the necessity of safeguarding these systems from insider-based data injection attacks, ensuring that collective decision-making remains effective and secure. Through rigorous theoretical analysis and algorithmic design, this part contributes novel methodologies that enhance the resilience and efficiency of decentralized learning frameworks.Part II shifts focus to the imperative of privacy preservation within energy systems, a domain where the collection and analysis of data have become indispensable for operational efficiency and innovation. With the advent of smart grids and distributed energy resources, the volume of sensitive data has surged, raising significant privacy concerns. This section proposes the use of differential privacy (DP) as a mathematically rigorous solution to protect individual data records while allowing for the aggregate analysis necessary for system optimization and planning. By applying DP mechanisms, the dissertation showcases how energy systems can share and utilize data for operational and research purposes without compromising individual privacy or data integrity. This part offers a comprehensive approach to developing privacy-preserving mechanisms, illustrating their application through clustering, synthetic data generation, and anomaly detection, thus facilitating secure and efficient data sharing in the energy sector.Together, these parts provide a holistic view of the challenges and solutions at the intersection of decentralized optimization and privacy preservation in networked systems. The dissertation contributes to the theoretical foundation and practical implementation of robust and privacy-preserving frameworks, offering significant insights for securing and optimizing decentralized systems in adversarial and privacy-sensitive contexts.
- 일반주제명
- Electrical engineering
- 일반주제명
- Computer science
- 일반주제명
- Statistics
- 일반주제명
- Information science
- 키워드
- Energy systems
- 기타저자
- Cornell University Electrical and Computer Engineering
- 기본자료저록
- Dissertations Abstracts International. 85-12A.
- 전자적 위치 및 접속
- 로그인 후 원문을 볼 수 있습니다.
MARC
008250123s2024 us c eng d■001000017161378
■00520250211151348
■006m o d
■007cr#unu||||||||
■020 ▼a9798382843346
■035 ▼a(MiAaPQ)AAI31242893
■040 ▼aMiAaPQ▼cMiAaPQ
■0820 ▼a621.3
■1001 ▼aRavi, Nikhil.▼0(orcid)0000-0002-1267-2563
■24510▼aEnhancing Networked Systems: A Comprehensive Approach to Robust and Privacy-Preserving Optimization Algorithms
■260 ▼a[Sl]▼bCornell University▼c2024
■260 1▼aAnn Arbor▼bProQuest Dissertations & Theses▼c2024
■300 ▼a205 p
■500 ▼aSource: Dissertations Abstracts International, Volume: 85-12, Section: A.
■500 ▼aAdvisor: Scaglione, Anna.
■5021 ▼aThesis (Ph.D.)--Cornell University, 2024.
■520 ▼aThis dissertation explores two pivotal aspects of networked systems operating in increasingly complex and adversarial environments: robust decentralized optimization and privacy preservation. Part I delves into the development of decentralized optimization algorithms, addressing the critical challenge of ensuring system robustness against adversarial threats. It investigates strategies to secure the integrity of consensus and optimization processes in distributed networks, where each node or agent contributes to a global objective without centralized coordination. This exploration underscores the necessity of safeguarding these systems from insider-based data injection attacks, ensuring that collective decision-making remains effective and secure. Through rigorous theoretical analysis and algorithmic design, this part contributes novel methodologies that enhance the resilience and efficiency of decentralized learning frameworks.Part II shifts focus to the imperative of privacy preservation within energy systems, a domain where the collection and analysis of data have become indispensable for operational efficiency and innovation. With the advent of smart grids and distributed energy resources, the volume of sensitive data has surged, raising significant privacy concerns. This section proposes the use of differential privacy (DP) as a mathematically rigorous solution to protect individual data records while allowing for the aggregate analysis necessary for system optimization and planning. By applying DP mechanisms, the dissertation showcases how energy systems can share and utilize data for operational and research purposes without compromising individual privacy or data integrity. This part offers a comprehensive approach to developing privacy-preserving mechanisms, illustrating their application through clustering, synthetic data generation, and anomaly detection, thus facilitating secure and efficient data sharing in the energy sector.Together, these parts provide a holistic view of the challenges and solutions at the intersection of decentralized optimization and privacy preservation in networked systems. The dissertation contributes to the theoretical foundation and practical implementation of robust and privacy-preserving frameworks, offering significant insights for securing and optimizing decentralized systems in adversarial and privacy-sensitive contexts.
■590 ▼aSchool code: 0058.
■650 4▼aElectrical engineering
■650 4▼aComputer science
■650 4▼aStatistics
■650 4▼aInformation science
■653 ▼aAdversarial resilience
■653 ▼aDecentralized optimization
■653 ▼aDifferential privacy
■653 ▼aEnergy systems
■653 ▼aPrivacy preservation
■690 ▼a0544
■690 ▼a0984
■690 ▼a0463
■690 ▼a0723
■690 ▼a0800
■71020▼aCornell University▼bElectrical and Computer Engineering.
■7730 ▼tDissertations Abstracts International▼g85-12A.
■790 ▼a0058
■791 ▼aPh.D.
■792 ▼a2024
■793 ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17161378▼nKERIS▼z이 자료의 원문은 한국교육학술정보원에서 제공합니다.
