메뉴 건너뛰기
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)
고객센터 ENG
주제분류

논문 기본 정보

저자정보
(한국과학기술원) (한국과학기술원) (한과학기술원) (한국과학기술원)
저널정보
대한전자공학회 대한전자공학회 학술대회 2024년도 대한전자공학회 하계학술대회 논문집
오류 신고하기

피인용 0

검색

    초록·키워드

    By Machine learning is deeply integrated into modern society, delivering outstanding performance in everyday life. However, these models still rely on the distribution of training data, which exacerbates data bias issues. As machine learning models aim to learn data distribution, biased training data significantly affects fairness and effectiveness. This study proposes a new methodology using adversarial training to develop classification models resilient to data bias. Although various attempts have been made to create unbiased classification models, they often rely on constructing excessively biased models or make unrealistic assumptions, such as knowing bias attribute labels. We aim to address these challenges through simple adversarial training. Experimentson various biased datasets showed that the proposed model training method reduces bias while maintaining accuracy. Furthermore, the model achieves high performance on unbiased datasets, demonstrating robust performance regardless of data distribution

    최근 본 자료 전체보기