http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
벼 등숙기 기온 및 수발아가 종실 품질 및 이화학적 특성에 미치는 영향
이현석,이윤호,황운하,정재혁,양서영,이충근,최명구,Lee, HyeonSeok,Lee, YunHo,Hwang, WoonHa,Jeong, JaeHyeok,Yang, SeoYeong,Lee, ChungGen,Choi, MyoungGoo 한국작물학회 2020 한국작물학회지 Vol.65 No.4
Studies on the occurrence of rice preharvest sprouting (PHS) have primarily focused on temperature and rainfall duration at the time of PHS induction, but average temperature during grain filling can have a great influence on PHS. This study analyzed the effect of average temperature during grain filling on PHS occurrence and subsequent changes in grain quality after PHS. For two consecutive years, average temperature differences during grain filling were produced by varying the transplanting date. Artificial rainfall was treated under identical accumulated temperatures of 1200℃ after heading. It was confirmed that the occurrence of PHS was higher under high average temperature conditions during grain filling. In addition, the degree of grain quality reduction caused by PHS occurred more severely under high temperature conditions during grain filling. In order to reduce the risk of PHS occurrence and subsequent quality damage, it is important to control the planting date to avoid high-temperature conditions during grain filling.
자가주의집중 기반의 심층 신경망을 이용한 Branchpoint 예측
이현석(Hyeonseok Lee),김성찬(Sungchan Kim) 한국정보과학회 2020 정보과학회논문지 Vol.47 No.4
접합(splicing)은 단백질 생성에 중요한 messenger RNA를 만들기 위한 RNA 처리 과정이다. Branchpoint는 접합 과정에서 중요한 역할을 하는 RNA 염기 서열 중 한 요소이며, 본 논문에서는 branchpoint 예측을 위한 심층 신경망 모델을 제안한다. Branchpoint 위치 예측에는 몇 가지 어려움이 존재한다. Branchpoint의 위치는 염기 서열에 존재하는 모티프(염기 서열에 나타나는 짧은 길이의 패턴)들과 밀접하게 관련된 것으로 알려져 있으며, RNA 서열의 특정 영역에 편중되어 있어 기계학습을 적용할 때 클래스불균형 문제가 발생한다. 본 논문의 기법은 입력 서열의 특징 학습을 위한 주의집중 기법과 클래스 불균형을 완화하기 위한 삼중항 손실을 활용한다. 이로 인해 알려진 최고 성능 수준의 예측 정확도를 유지하면서, branchpoint 위치에 영향을 주는 모티프들을 찾아내 예측에 대한 설명을 가능하게 해준다. Splicing is a ribonucleic acid (RNA) process of creating a messenger RNA (mRNA) translated into proteins. Branchpoints are sequence elements of RNAs essential in splicing. This paper proposes a novel method for branchpoint prediction. Identification of branchpoints involves several challenges. Branchpoint sites are known to depend on several sequence patterns, called motifs. Also, a branchpoint distribution is highly biased, imposing a class-imbalanced problem. Existing approaches are limited in that they either rely on handcrafted sequential features or ignore the class imbalance. To address those difficulties, the proposed method incorporates 1) Attention mechanisms to learn sequence-positional long-term dependencies, and 2) Regularization with triplet loss to alleviate the class imbalance. Our method is comparable to the state-of-the-art performance while providing rich interpretability on its decisions.
벼 오존가스 노출에 따른 초기 생육 및 안토시아닌 생합성 변화 분석 연구
이현석 ( Hyeonseok Lee ),황운하 ( Woonha Hwang ),양서영 ( Seoyeong Yang ),송영서 ( Yeongseo Song ),임우진 ( Woojin Im ),정회정 ( Hoejeong Jeong ),이충근 ( Chunggen Lee ),김주희 ( Juhee Kim ),최명구 ( Myounggoo Choi ) 한국농림기상학회 2023 한국농림기상학회지 Vol.25 No.2
Recently, the concentration of fine dustcausative substances (NOx, VOC, etc.) in the atmosphere has increased, resulting in high concentrations of tropospheric ozone (O3) and increased damage to crops. This study aimed to analyze the impact of high concentrations of ozone gas on the initial growth of rice plants and investigate the relationship between ozone damage resistance and anthocyanin biosynthesis. To achieve this, rice plants were exposed to elevated levels of ozone gas using an ozone chamber, and subsequent measurements were taken to assess changes in growth, the percentage of damaged leaves, and the anthocyanin content. The results revealed that varieties with a higher proportion of damaged leaves exhibited a relative increase in anthocyanin biosynthesis following ozone exposure. Notably, detrimental effects on growth, such as decreased biomass, were mitigated. Additionally, Anthocyanin biosynthesis genes in rice were listed by selecting homologous genes from Arabidopsis and Maize. The expression of OsF3H2, OsFLS1 and OsLDOX3 was induced during ozone treatment. This result is expected to contribute to the study of the protection mechanism of plants from ozone damage.
신경망 근사에 의한 다중 레이어의 클래스 활성화 맵을 이용한 블랙박스 모델의 시각적 설명 기법
강준규,전민경,이현석,김성찬,Kang, JuneGyu,Jeon, MinGyeong,Lee, HyeonSeok,Kim, Sungchan 대한임베디드공학회 2021 대한임베디드공학회논문지 Vol.16 No.4
In this paper, we propose a novel visualization technique to explain the predictions of deep neural networks. We use knowledge distillation (KD) to identify the interior of a black-box model for which we know only inputs and outputs. The information of the black box model will be transferred to a white box model that we aim to create through the KD. The white box model will learn the representation of the black-box model. Second, the white-box model generates attention maps for each of its layers using Grad-CAM. Then we combine the attention maps of different layers using the pixel-wise summation to generate a final saliency map that contains information from all layers of the model. The experiments show that the proposed technique found important layers and explained which part of the input is important. Saliency maps generated by the proposed technique performed better than those of Grad-CAM in deletion game.