• 한국수자원학회논문집
• /
• 제51권4호
• /
• pp.281-291
• /
• 2018

화음탐색법은 2001년 고려대학교 수자원연구실에서 개발한 최적화 알고리즘으로 재즈의 즉흥연주에서 반복적인 연습을 거듭할 수 록 좋은 화음이 만들어지는 현상에 착안하였다. 화음탐색법은 처음 소개된 논문이 Google Scholar 기준 약 3,600여 회(2018년 1월 11일 기준) 인용될 만큼 유전자알고리즘과 견줄만한 세계적인 최적화 알고리즘이 되었고 비단 수자원공학 및 토목공학 뿐 만 아니라 공학 전 분야, 의학, 경영학, 인문학 등 다양한 분야에 적용되고 있다. 본 논문은 화음탐색법을 포함한 최적화 알고리즘이 수자원공학의 다양한 분야에서 널리 적용되기를 바라며 작성된 화음탐색법 총설논문(Review Article)이다. 따라서, 본 논문에서는 먼저 화음탐색법을 간략히 소개하고 적용분야 및 분야별 적용 빈도를 살펴본다. 또한 화음탐색법의 세계화 현황을 관련 학회의 성장과 관련 연구프로젝트의 동향 정리를 통해 알아본다. 마지막으로 국내 수자원공학 분야 연구에 적용된 최적화 알고리즘 현황을 살펴보고 활용의 증대를 위한 몇 가지 제안사항을 전달하며 마무리한다.

• 패션비즈니스
• /
• 제8권1호
• /
• pp.34-48
• /
• 2004

The purpose of this study is to explore emotion aspects of consumers purchasing clothing through cable television home shopping. Qualitative research method is used to widely understand how emotion aspects of consumers have effected on their purchasing behavior. The results of depth interviews may be classified into 13 feelings factors satisfaction, pleasure/delight, respect, attraction, fresh, convenience, unburdened, emptiness, displeasure/temper, anxiety, tedious, distrust, regret. The content of information acquiring from the process of clothing purchase decision making is analysed. In the problem recognition stage, purchase motivation were physical space (around people) and imaginary space(by how clothing goods are introduced to consumers thorough TV monitor). In the information search stage, purchasing action patterns to search information were situational pattern and habitual pattern. In alternative evaluation stage, the considering best important factors to choice clothes were quality, price, design, and color. In purchase stage, consumers said they felt anxiety, because of characteristics of purchase way that they should pay first and then received the ordered goods a fews days later. In post-purchase behavior stage, if consumers satisfied goods purchased through TV home shopping, they recommended it to around others, but unsatisfied with ordered goods, they tried to refund, exchange with anther one, or write it on homepage of the home shopping company.

• 대한전기학회논문지:전력기술부문A
• /
• 제53권11호
• /
• pp.596-603
• /
• 2004

This paper proposes an expert system based on the pattern recognition method which can enhance the accuracy and effectiveness of real-time bus reconfiguration strategy for the transfer of faulted load when a main transformer fault occurs in the automated substation. The minimum distance classification method is adopted as the pattern recognition method of expert system. The training pattern set is designed MTr by MTr to minimize the searching time for target load pattern which is similar to the real-time load pattern. But the control pattern set, which is required to determine the corresponding bus reconfiguration strategy to these trained load pattern set is designed as one table by considering the efficiency of knowledge base design because its size is small. The training load pattern generator based on load level and the training load pattern generator based on load profile are designed, which are can reduce the size of each training pattern set from max L/sup (m+f)/ to the size of effective level. Here, L is the number of load level, m and f are the number of main transformers and the number of feeders. The one reduces the number of trained load pattern by setting the sawmiller patterns to a same pattern, the other reduces by considering only load pattern while the given period. And control pattern generator based on exhaustive search method with breadth-limit is designed, which generates the corresponding bus reconfiguration strategy to these trained load pattern set. The inference engine of the expert system and the substation database and knowledge base is implemented in MFC function of Visual C++ Finally, the performance and effectiveness of the proposed expert system is verified by comparing the best-first search solution and pattern recognition solution based on diversity event simulations for typical distribution substation.

• 대한전기학회논문지:전력기술부문A
• /
• 제53권12호
• /
• pp.637-644
• /
• 2004

This paper proposes an expert system with the knowledge learning capability which can enhance the safety and effectiveness of substation operation in the automated substation as well as existing substation by inferring multiple events such as main transformer fault, busbar fault and main transformer work schedule under multiple inference mode and multiple objective mode and by considering totally the switch status and the main transformer operating constraints. Especially inference mode includes the local minimum tree search method and pattern recognition method to enhance the performance of real-time bus reconfiguration strategy. The inference engine of the expert system consists of intuitive inferencing part and logical inferencing part. The intuitive inferencing part offers the control strategy corresponding to the event which is most similar to the real event by searching based on a minimum distance classification method of pattern recognition methods. On the other hand, logical inferencing part makes real-time control strategy using real-time mode(best-first search method) when the intuitive inferencing is failed. Also, it builds up a knowledge base or appends a new knowledge to the knowledge base using pattern learning function. The expert system has main transformer fault, main transformer maintenance work and bus fault processing function. It is implemented as computer language, Visual C++ which has a dynamic programming function for implementing of inference engine and a MFC function for implementing of MMI. Finally, it's accuracy and effectiveness is proved by several event simulation works for a typical substation.

• Journal of Cardiovascular Imaging
• /
• 제31권4호
• /
• pp.159-168
• /
• 2023

BACKGROUND: The gold standard for diagnosis of cardiac tumours is histopathological examination. Cardiovascular magnetic resonance (CMR) is a valuable non-invasive, radiation-free tool for identifying and characterizing cardiac tumours. Our aim is to understand CMR diagnosis of cardiac tumours by distinguishing benign vs. malignant tumours compared to the gold standard. METHODS: A systematic search was performed in the PubMed, Web of Science, and Scopus databases up to December 2022, and the results were reviewed by 2 independent investigators. Studies reporting CMR diagnosis were included in a meta-analysis, and pooled measures were obtained. The risk of bias was assessed using the Quality Assessment Tools from the National Institutes of Health. RESULTS: A total of 2,321 results was obtained; 10 studies were eligible, including one identified by citation search. Eight studies were included in the meta-analysis, which presented a pooled sensitivity of 93% and specificity of 94%, a diagnostic odds ratio of 185, and an area under the curve of 0.98 for CMR diagnosis of benign vs. malignant tumours. Additionally, 4 studies evaluated whether CMR diagnosis of cardiac tumours matched specific histopathological subtypes, with 73.6% achieving the correct diagnosis. CONCLUSIONS: To the best of our knowledge, this is the first published systematic review on CMR diagnosis of cardiac tumours. Compared to histopathological results, the ability to discriminate benign from malignant tumours was good but not outstanding. However, significant heterogeneity may have had an impact on our findings.

• 지능정보연구
• /
• 제27권1호
• /
• pp.103-128
• /
• 2021

최근 주식의 수익률과 거래량을 설명하는 주요 요인으로서 투자자의 관심도와 주식 관련 정보 전파의 영향력이 부각되고 있다. 또한 인공지능과 같은 혁신 신기술을 개발보급하거나 활용하려는 기업의 경우 거시환경 및 시장 불확실성 때문에 기업의 미래 주식 수익률과 주식 변동성을 예측하기 어렵다는 문제를 가지고 있다. 이는 인공지능 활성화의 장애요인으로 인식되고 있다. 따라서 본 연구의 목적은 인공지능 관련 기술 키워드의 인터넷 검색량을 투자자의 관심 척도로 사용하여, 기업의 주가 변동성을 예측하는 기계학습 모형을 제안하는 것이다. 이를 위해 심층신경망 LSTM(Long Short-Term Memory)과 벡터자기회귀(Vector Autoregression)를 통해 주식시장을 예측하고, 기술의 사회적 수용 단계에 따라 키워드 검색량을 활용한 주가예측 성능 비교를 통해 기업의 투자수익 예측이나 투자자들의 투자전략 의사결정을 지원하는 주가 예측 모형을 구축하였다. 또한 인공지능 기술의 세부 하위 기술에 대한 분석도 실시하여 기술 수용 단계에 따른 세부 기술 키워드 검색량의 변화를 살펴보고 세부기술에 대한 관심도가 주식시장 예측에 미치는 영향을 살펴보았다. 이를 위해 본 연구에서는 인공지능, 딥러닝, 머신러닝 키워드를 선정하여, 2015년 1월 1일부터 2019년 12월 31일까지 5년간의 인터넷 주별 검색량 데이터와 코스닥 상장 기업의 주가 및 거래량 데이터를 수집하여 분석에 활용하였다. 분석 결과 인공지능 기술에 대한 키워드 검색량은 사회적 수용 단계가 진행될수록 증가하는 것으로 나타났고, 기술 키워드를 기반으로 주가예측을 하였을 경우 인식(Awareness)단계에서 가장 높은 정확도를 보였으며, 키워드별로 가장 좋은 예측 성능을 보이는 수용 단계가 다르게 나타남을 확인하였다. 따라서 기술 키워드를 활용한 주가 예측 모델 구축을 위해서는 해당 기술의 하위 기술 분류를 고려할 필요가 있다. 본 연구의 결과는 혁신기술을 기반으로 기업의 투자수익률을 예측하기 위해서는 기술에 대한 대중의 관심이 급증하는 인식 단계를 포착하는 것이 중요하다는 점을 시사한다. 또한 최근 금융권에서 선보이고 있는 빅데이터 기반 로보어드바이저(Robo-advisor) 등 투자 의사 결정 지원 시스템 개발 시 기술의 사회적 수용도를 세분화하여 키워드 검색량 변화를 통해 예측 모델의 정확도를 개선할 수 있다는 점을 시사하고 있다.

• 한국콘텐츠학회논문지
• /
• 제7권6호
• /
• pp.22-30
• /
• 2007

대부분의 실세계의 시스템들, 즉 경제, 주식시장, 의료분야 등의 많은 시스템들은 동적이며 복잡한 현상을 갖는다. 이러한 특징들의 시스템을 이해하는 전형적인 방법은 시스템행위에 대한 모델을 세우고 분석하는 것이다. 본 연구에서는 실세계의 동적 시스템에서 발생되는 시계열데이터들에 대하여 최적의 클러스터를 형성하기 위한 방법을 연구한다. 먼저 클러스터 수를 결정하는 기준으로 베이지안정보기준(BIC : Bayesian Information Criterion)근사법의 활용도를 검증하고 데이터 크기와 베이지안정보기준값의 상관관계를 파악함으로 탐색 효율을 높이는 방안을 제안하며 클러스터링 과정으로 모델기반과 유사기반의 방법론을 비교 확인하여 본다. 실제의 시계열데이터(주가)에 대해 실험을 시행하였고 베이지안정보기준 근사 측도는 데이터의 크기에 따라 파티션의 사이즈를 정확히 추정하는 것을 확인하였으며 또한 유사기반의 방식보다 모델기반의 방법론이 클러스터링에서 더 나은 결과를 갖는 것을 확인하였다.

• Archives of Pharmacal Research
• /
• 제27권4호
• /
• pp.381-385
• /
• 2004

Nine phenolic ($1\~9$) and two furan type (10, 11) compounds, were isolated from the methanolic extract of the tuber of Gastrodia elata Blume (Orchidaceae) in the course of continuing search for platelet anit-aggregating plant components. Compound 1 was identified as 4,4'-dihy-droxybenzyl sulfone, a novel compound for the best of our knowledge. Compound 10, 5-hydroxymethyl-2-furancarboxaldehyde, was isolated for the first time from this plant. Compound 1 ($IC_{50};\;83{\mu}M$) was about four times more inhibitory to U46619 induced aggregation than ASA ($IC_{50};\340{\mu}M$). Compound 9, 4,4'-dihydroxy-dibenzylether, ($IC_{50};\;5{\mu}M$, $3{\mu}M\;and\;33{\mu}M$, respectively) was $10\~}80$ fold more potent than ASA ($IC_{50};\;420\;{\mu}M,\;53\;{\mu}M\;and\;340\;{\mu}M$ respectively) to collagen, epinephrine and U46619 induced aggregation, although it is less active than ASA to AA induced aggregation.

• Journal of the Korean Society for Industrial and Applied Mathematics
• /
• 제12권4호
• /
• pp.227-238
• /
• 2008

In this paper we extend primal-dual interior point algorithm for linear optimization (LO) problems to $P_*({\kappa})$ linear complementarity problems(LCPs) ([1]). We define proximity functions and search directions based on kernel functions, ${\psi}(t)=\frac{t^{p+1}-1}{p+1}-{\log}\;t$, $p{\in}$[0, 1], which is a generalized form of the one in [16]. It is the first to use this class of kernel functions in the complexity analysis of interior point method(IPM) for $P_*({\kappa})$ LCPs. We show that if a strictly feasible starting point is available, then new large-update primal-dual interior point algorithms for $P_*({\kappa})$ LCPs have $O((1+2{\kappa})nlog{\frac{n}{\varepsilon}})$ complexity which is similar to the one in [16]. For small-update methods, we have $O((1+2{\kappa})\sqrt{n}{\log}{\frac{n}{\varepsilon}})$ which is the best known complexity so far.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제8권2호
• /
• pp.678-690
• /
• 2014

The current networks do not disclose information about a management domain due to scalability, manageability and commercial reasons. Therefore, it is very hard to calculate an optimal path to the destination. Also, due to poor information sharing, if an error occurs in the intermediate path, it is very difficult to re-search the path and find the best path. Hence, to manage each domain more efficiently, an architecture with top-level path computation node which can obtain information of separate nodes are highly needed This study aims to investigate a federation of a united network around NRN(National Research Network) that could allow resource sharing between countries and also independent resource management for each country. Considering first the aspects that can be accessed from the perspective of a national research network, ICE(Information Control Element) and GFO(Global Federation Organizer)-based architecture is designed as a top-level path computation element to support traffic engineering and applied to the multi-domain network. Then, the federation for the independent management of resources and resource information sharing among national research networks have been examined.