메뉴 건너뛰기
.. 내서재 .. 알림
소속 기관/학교 인증
인증하면 논문, 학술자료 등을  무료로 열람할 수 있어요.
한국대학교, 누리자동차, 시립도서관 등 나의 기관을 확인해보세요
(국내 대학 90% 이상 구독 중)
로그인 회원가입 고객센터 ENG
주제분류

추천
검색
질문

논문 기본 정보

자료유형
학술저널
저자정보
Yu-Nam Cheong (Sunchon National University) Kwang-Seong Shin (Sunchon National University) Seong-Yoon Shin (Kunsan National University)
저널정보
한국정보통신학회JICCE Journal of information and communication convergence engineering Journal of information and communication convergence engineering Vol.23 No.1
발행연도
2025.3
수록면
1 - 7 (7page)

이용수

표지
📌
연구주제
📖
연구배경
🔬
연구방법
🏆
연구결과
AI에게 요청하기
추천
검색
질문

초록· 키워드

오류제보하기
Although federated learning (FL) has emerged as a revolutionary paradigm for privacy-preserving distributed learning, the non-independent and identically distributed nature of client data poses significant challenges to model performance. This study proposes a personalized FL framework that integrates dynamic client sampling and advanced personalization techniques. The proposed method effectively addresses data heterogeneity, enhances client-specific performance, and ensures efficient communication. Experimental results on CIFAR-100 demonstrate significant improvements, with personalized models reaching accuracies between 53% and 57% compared to 23.13% for the global model, while maintaining stable communication costs of 376,588 bytes per round. Furthermore, our framework demonstrates robust performance across different levels of data heterogeneity, maintaining consistent accuracy even when the Gini coefficient of the label distribution varies from 0.2 to 0.8. Statistical analysis confirms the significance of these improvements, demonstrating the effectiveness of FedPer in handling heterogeneous data distributions while maintaining computational efficiency, making the proposed method particularly suitable for resource-constrained, privacy-sensitive applications.

목차

Abstract
I. INTRODUCTION
II. RELATED WORKS
III. METHODOLOGY
IV. EXPERIMENTAL RESULTS
V. CONCLUSION
REFERENCES

참고문헌 (0)

참고문헌 신청

함께 읽어보면 좋을 논문

논문 유사도에 따라 DBpia 가 추천하는 논문입니다. 함께 보면 좋을 연관 논문을 확인해보세요!

이 논문의 저자 정보

최근 본 자료

전체보기

댓글(0)

0