http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Classification of impinging jet flames using convolutional neural network with transfer learning
Minwoo Lee,Sangwoong Yoon,Ju Han Kim,Yuangang Wang,Kee-Man Lee,Frank Chongwoo Park,Chae Hoon Sohn 대한기계학회 2022 JOURNAL OF MECHANICAL SCIENCE AND TECHNOLOGY Vol.36 No.3
Depending on the equivalence ratio and the Reynolds number, impinging jet flames exhibit several modes of thermoacoustic oscillation. In this study, we present a machine-learning-based method for classifying the regimes of thermoacoustic oscillation. We perform transfer learning to train the convolutional neural network model designed to classify flame images. We show that an accurate classification of impinging jet flames is achieved with an accuracy of 93.6 % by using just a single snapshot image. This study constitutes the first demonstration of transfer learning in classifying fluid images, opening up new possibilities for robust image-based diagnostics of various fluid and combustion systems.
Jung, Minwoo,Yoon, Youngwoon,Park, Jae Hoon,Cha, Wonsuk,Kim, Ajeong,Kang, Jinback,Gautam, Sanjeev,Seo, Dongkyun,Cho, Jeong Ho,Kim, Hyunjung,Choi, Jong Yong,Chae, Keun Hwa,Kwak, Kyungwon,Son, Hae Jung American Chemical Society 2014 ACS NANO Vol.8 No.6
<P>We synthesized a series of acceptor–donor–acceptor-type small molecules (SIDPP-EE, SIDPP-EO, SIDPP-OE, and SIDPP-OO) consisting of a dithienosilole (SI) electron-donating moiety and two diketopyrrolopyrrole (DPP) electron-withdrawing moieties each bearing linear <I>n</I>-octyl (O) and/or branched 2-ethylhexyl (E) alkyl side chains. X-ray diffraction patterns revealed that SIDPP-EE and SIDPP-EO films were highly crystalline with pronounced edge-on orientation, whereas SIDPP-OE and SIDPP-OO films were less crystalline with a radial distribution of molecular orientations. Near-edge X-ray absorption fine structure spectroscopy disclosed an edge-on orientation with a molecular backbone tilt angle of ∼22° for both SIDPP-EE and SIDPP-EO. Our analysis of the molecular packing and orientation indicated that the shorter 2-ethylhexyl groups on the SI core promote tight π–π stacking of the molecular backbone, whereas <I>n</I>-octyl groups on the SI core hinder close π–π stacking to some degree. Conversely, the longer linear <I>n</I>-octyl groups on the DPP arms facilitate close intermolecular packing <I>via</I> octyl–octyl interdigitation. Quantum mechanics/molecular mechanics molecular dynamics simulations determined the optimal three-dimensional positions of the flexible alkyl side chains of the SI and DPP units, which elucidates the structural cause of the molecular packing and orientation explicitly. The alkyl-chain-dependent molecular stacking significantly affected the electrical properties of the molecular films. The edge-on oriented molecules showed high hole mobilities in organic field-effect transistors, while the radially oriented molecules exhibited high photovoltaic properties in organic photovoltaic cells. These results demonstrate that appropriate positioning of alkyl side chains can modulate crystallinity and molecular orientation in SIDPP films, which ultimately have a profound impact on carrier transport and photovoltaic performance.</P><P><B>Graphic Abstract</B> <IMG SRC='http://pubs.acs.org/appl/literatum/publisher/achs/journals/content/ancac3/2014/ancac3.2014.8.issue-6/nn501133y/production/images/medium/nn-2014-01133y_0010.gif'></P><P><A href='http://pubs.acs.org/doi/suppl/10.1021/nn501133y'>ACS Electronic Supporting Info</A></P>
Kim, Minwoo,Kim, Hyun Ryoung,Chae, Su Young,Larson, Ronald G.,Lee, Hwankyu,Park, Jae Chan American Chemical Society 2013 The journal of physical chemistry. B, Condensed ma Vol.117 No.23
<P>Heparin decomplexation experiments, as well as all-atom (AA) and coarse-grained (CG) molecular dynamics (MD) simulations, were performed to determine the effect of the size of arginine(Arg)-rich peptides on the structure and binding strength of the siRNA–peptide complex. At a fixed peptide/siRNA mole ratio of 5:1 or 10:1, the siRNA complexes with peptides longer than nine Arg residues are more easily decomplexed by heparin than are those with nine Arg residues. At these mole ratios, peptides longer than nine Arg residues have cationic/anionic charge ratios in excess of unity, and produce more weakly bound complexes than nine Arg residue ones do. AA simulations of mixtures of peptides with a single siRNA show formation of an electrostatically induced complex, and the longer peptides produce a larger complex, but with no significant increase in the number of Arg residues bound to the siRNA. Larger-scale CG-MD simulations show that multiple siRNAs can be linked together by peptides into a large complex, as observed in the experiments. The peptides longer than nine residues, which at mole ratio 5:1 yield a peptide/siRNA charge ratio in excess of unity, include many noninteracting Arg residues, which repel each other electrostatically. This leads to a less dense complex than for 9-residue peptides, which can explain why these longer complexes are more easily decomplexed by heparin molecules, as observed in the experiments. The key role of the charge ratio is supported by simulations that show that, at a mole ratio of 2.5 peptides per siRNA, the longer 18-residue peptide has a charge ratio of roughly unity and also shows a tight complex, just as the 9-residue peptide does at a 5:1 mole ratio, where its charge ratio is also unity.</P><P><B>Graphic Abstract</B> <IMG SRC='http://pubs.acs.org/appl/literatum/publisher/achs/journals/content/jpcbfk/2013/jpcbfk.2013.117.issue-23/jp402868g/production/images/medium/jp-2013-02868g_0014.gif'></P><P><A href='http://pubs.acs.org/doi/suppl/10.1021/jp402868g'>ACS Electronic Supporting Info</A></P>
Additive time-dependent hazard model with doubly truncated data
Gordon Frank,Minwoo Chae,김용대 한국통계학회 2019 Journal of the Korean Statistical Society Vol.48 No.2
For doubly truncated data, i.e. the variables of interest are only observable if they lie in a certain random interval, an additive hazard model with time-dependent regression coefficients is investigated. Consistency and asymptotic normality are proven under mild assumptions. A simulation study investigates the finite sample properties and the influence of the truncation distribution on the estimation error. Finally, the method is applied to a doubly truncated data set of German companies, where the age at insolvency is of interest.
채민우,강민수,김용대,Chae, Minwoo,Kang, Minsoo,Kim, Yongdai 한국데이터정보과학회 2013 한국데이터정보과학회지 Vol.24 No.5
본 연구에서는 논문이나 특허 등의 문서들의 인용 정보를 활용하여 연관성이 높고 중요한 특허를 추천하는 방법을 제안한다. 문서 간의 연관성 지표인 공통피인용횟수와 중요도 지표인 HITS를 적절한 형태로 결합한 뉴먼 커널로부터 두 정보의 반영 정도를 조율하는 것이 핵심이다. 제안하는 방법은 미래의 인용에 대한 예측 오차를 최소화하는 것으로 이를 통해 뉴먼 커널의 조율모수 ${\gamma}$를 적절하게 선택할 수 있다. 또한, 거대 인용 자료를 분석하기 위해 필요한 계산 기술에 대해서 자세히 논의한다. 마지막으로, 미국 등록 특허 400만 건에 대한 실증적 자료 분석을 시행한다. In this research, we propose a document recommendation method which can find documents that are relatively important to a specific document based on citation information. The key idea is parameter tuning in the Neumann kernal which is an intermediate between a measure of importance (HITS) and of relatedness (co-citation). Our method properly selects the tuning parameter ${\gamma}$ in the Neumann kernal minimizing the prediction error in future citation. We also discuss some comutational issues needed for analysing large citation data. Finally, results of analyzing patents data from the US Patent Office are given.
김용대,최민우,Kim, Yongdai,Chae, Minwoo 한국통계학회 2014 응용통계연구 Vol.27 No.6
Hjort (1990)가 제안한 베타과정은 베이지안 생존분석 또는 사건사 분석에서 널리 쓰이는 사전분포이다. 본 논문은 베타과정에 대한 최신 이론과 이를 기반으로 하는 베이지안 생존자료분석 방법을 주로 다룬다. 구체적으로는 베타과정의 생성법, 사후 분포, 대표본 이론, 베이지안 계산법, 혼합베타과정 등을 소개하기로 한다. This article is concerned with one of the most important prior distributions for Bayesian analysis of survival and event history data, called Beta processes, proposed in Hjort (1990). We review the current state of the art of beta processes and their application to survival analysis. Relevant methodological and practical areas of research that we touch on relate to constructions, posterior distributions, large-sample properties, Bayesian computations, and mixtures of Beta processes.
SOC Balance Strategy of Li-ion Battery System by Using Fuzzy Logic Control System
Prakash Thapa(쁘러카스 타파),Jinhyeok Kim(김진혁),Minwoo Lee(이민우),hyunjun Ji(지현준),Sooyeon Chae(채수연),Wolguk Park(박월국),Seunghyun An(안승현),Sung Gi Kwon(권성기),Kil Ju Na(나길주),Gye Choon Park(박계춘),Jin Lee(이진) 한국신재생에너지학회 2018 한국신재생에너지학회 학술대회논문집 Vol.2018 No.11