http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
클래스별 규제 엔트로피를 이용한 시각적으로 유사한 객체의 분류
김예찬(Yechan Kim),이윤관(Younkwan Lee),고영민(YeongMin Ko),유형준(Hyeongjun Yoo),전병관(Byung Gwan Jeon),전문구(Moongu Jeon) 한국정보과학회 2021 한국정보과학회 학술발표논문집 Vol.2021 No.6
본 논문은 신경망이 시각적으로 유사한 객체를 더욱 강건하게 분류할 수 있도록 클래스별 규제 엔트로피를 채택한 새로운 손실 함수를 제안한다. 이러한 객체는 종종 신경망 모델의 예측 신뢰도를 저하시킨다. 비록 교차 엔트로피 (Cross Entropy)로 학습된 모델은 일반적으로 잘 동작하지만, 교차 엔트로피는 시각적으로 유사한 서로 다른 클래스를 분류하는데 있어 불충분하다. 이는 교차 엔트로피는 오직 Ground Truth (GT) 클래스의 예측 확률만을 이용하고, 그 외의 네거티브 클래스의 예측 확률을 모델의 학습을 위해 유의미하게 활용하지 못하기 때문이다. 이 한계점을 극복하기 위해, 본 연구는 GT가 아닌 클래스의 예측확률을 클래스별로 차등적으로 규제하는 새로운 신경망 훈련 방법을 제시한다. 모델의 각 클래스에 대한 예측 확률을 GT의예측을 위협하는 정도로 해석하는 새로운 관점으로부터 제안된 본 전략은 Ground Truth 클래스가 모델의 추론 과정에서 더욱 강건하게 선택되는 효과로 이어진다. 우리는 CIFAR, SVHN, Tiny-ImageNet에서 광범위한 실험을 수행하여 종래의 방법에 비해 제안하는 방법이 더 높은 분류 정확도를 달성함을 확인하였다.
25.8 kV 옥내용 노후 GIS O-ring의 건전도 평가에 관한 연구
김예찬(Yechan Kim),정민경(Minkyung Jeong),조우현(Uhyeon Jo),Ragil Handito,구본혁(Bonhyuk Ku),강형구(Hyoungku Kang) 대한전기학회 2024 전기학회논문지 Vol.73 No.5
Asset management in power industries is crucial for maximizing asset value by considering life-cycle data and ensuring reliability and cost-effectiveness. This paper addresses the paradigm shift in asset management strategies from traditional approaches to Condition-Based Risk Management (CBRM). Unlike international practices, South Korea relies on Risk-Based Replacement (RBR), which assesses risk based on equipments years of service, leading to premature replacements of Gas-Insulated Switchgear (GIS) without considering equipment condition. Consequently, asset value maximization is hindered. To address this, the paper evaluates the degradation characteristics and health condition of aging GIS O-rings used for approximately 27 years. The collected O-ring samples were in excellent condition. The findings contribute to efficient maintenance cycle determination and the establishment of economical and efficient power facility asset management strategies.
지능형 메디컬 기기 개발을 위한 KANO-QFD 모델 제안: AI 기반 탈모관리 기기 중심으로
김예찬(Yechan Kim),최광은(Kwangeun Choi),정두희(Doohee Chung) 한국지능정보시스템학회 2022 지능정보연구 Vol.28 No.1
With the launch of Artificial Intelligence(AI)-based intelligent products on the market, innovative changes are taking place not only in business but also in consumers daily lives. Intelligent products have the potential to realize technology differentiation and increase market competitiveness through advanced functions of artificial intelligence. However, there is no new product development methodology that can sufficiently reflect the characteristics of artificial intelligence for the purpose of developing intelligent products with high market acceptance. This study proposes a KANO-QFD integrated model as a methodology for intelligent product development. As a specific example of the empirical analysis, the types of consumer requirements for hair loss prediction and treatment device were classified, and the relative importance and priority of engineering characteristics were derived to suggest the direction of intelligent medical product development. As a result of a survey of 130 consumers, accurate prediction of future hair loss progress, future hair loss and improved future after treatment realized and viewed on a smartphone, sophisticated design, and treatment using laser and LED combined light energy were realized as attractive quality factors among the KANO categories. As a result of the analysis based on House of Quality of QFD, learning data for hair loss diagnosis and prediction, micro camera resolution for scalp scan, hair loss type classification model, customized personal account management, and hair loss progress diagnosis model were derived. This study is significant in that it presented directions for the development of artificial intelligence-based intelligent medical product that were not previously preceded.
25.8 kV 옥내용 노후 GIS O-ring의 건전도 평가에 관한 연구
김예찬(Yechan Kim),정민경(Minkyung Jeong),조우현(Uhyeon Jo),Ragil Handito,구본혁(Bonhyuk Ku),강형구(Hyoungku Kang) 대한전기학회 2024 전기학회논문지 Vol.73 No.5
Asset management in power industries is crucial for maximizing asset value by considering life-cycle data and ensuring reliability and cost-effectiveness. This paper addresses the paradigm shift in asset management strategies from traditional approaches to Condition-Based Risk Management (CBRM). Unlike international practices, South Korea relies on Risk-Based Replacement (RBR), which assesses risk based on equipments years of service, leading to premature replacements of Gas-Insulated Switchgear (GIS) without considering equipment condition. Consequently, asset value maximization is hindered. To address this, the paper evaluates the degradation characteristics and health condition of aging GIS O-rings used for approximately 27 years. The collected O-ring samples were in excellent condition. The findings contribute to efficient maintenance cycle determination and the establishment of economical and efficient power facility asset management strategies.