• Nuclear Engineering and Technology
• /
• 제47권2호
• /
• pp.176-186
• /
• 2015

Pattern classifications have become important tools for fault diagnosis in nuclear power plants (NPP). However, it is often difficult to obtain training data under fault conditions to train a supervised classification model. By contrast, normal plant operating data can be easily made available through increased deployment of supervisory, control, and data acquisition systems. Such data can also be used to train classification models to improve the performance of fault diagnosis scheme. In this paper, a fault diagnosis scheme based on semisupervised classification (SSC) scheme is developed. In this scheme, new measurements collected from the plant are integrated with data observed under fault conditions to train the SSC models. The trained models are subsequently applied to new measurements for fault diagnosis. In comparison with supervised classifiers, the proposed scheme requires significantly fewer data collected under fault conditions to train the classifier. The developed scheme has been validated using different fault scenarios on a desktop NPP simulator as well as on a physical NPP simulator using a graph-based SSC algorithm. All the considered faults have been successfully diagnosed. The results have demonstrated that SSC is a promising tool for fault diagnosis in NPPs.

• 교육시설 논문지
• /
• 제24권4호
• /
• pp.15-24
• /
• 2017

This paper aims to investigate in what extent subject-centered clusters are different from one another in terms of message system, which is composed of curriculum, pedagogy and evaluation. For this, Bernstein's pedagogic transmission code(i.e., classification and framing) and school typology(i.e., open-type or close-type) have been explored, and then applied into Shimin Junior School, Japan, in order to find out substantial characteristics between subject-centered clusters. In this case study, VGA(visibility graph analysis), as one of syntactical methodologies in space syntax theory, has been used to measure to what degree they are actually different. Throughout in-depth investigation of spatial configurations, it can be said that the square of clusters is strongly connected and integrated very well, so that it acts as an anchor place for school life within a cluster. However, it works in different ways according to message systems. In the subjects like Japanese and Science whose message system are characterized by strong classification and strong framing, integration values are relatively low, and this means that it is hard to expect cross-referencing activities through the subject squares. On the contrary, the subject of Social Studies defined by weak classification and weak framing shows the highest mean integration values, and this can be expected that there are inter-changeable learning activities in the square.

• 대한원격탐사학회지
• /
• 제22권6호
• /
• pp.613-619
• /
• 2006

본 연구는 원격 탐사의 영상 처리에서 영상 분할의 상위 수준으로 응집 계층 clustering의 dendrogram을 통한 무감독 영상 분류를 제안한다. 제안된 알고리즘은 분광 영역에서 정의된 RAG (Regional Agency Graph)와 min-heap 자료 구조를 이용하여 MCSNP (Mutual Closest Spectral Neighbor Pair)의 집합을 검색하면서 합병을 수행하는 계층 clustering 방법이다. 계산 시간과 저장 기억의 사용에 대한 효율을 증가시키기 위해 분광적 인접성을 정의하는 분광 공간(spectral space)내의 다중 창을 사용하였고 RNV (Region Neighbor Vector)을 이용하여 합병에 의하여 변하는 RAG 갱신하였고 적정한 단계 수가 주어진다면 제안된 알고리즘은 집단 합병의 계층적 관계를 쉽게 해석 할 수 있는 dendrogram을 생성한다. 본 연구는 simulation 자료를 사용하여 광범위하게 제안된 알고리즘에 대한 평가 실험을 수행 하였으며 실험 결과는 알고리즘의 효율성을 입증하였다. 또한 한반도에서 관측된 방대한 크기의 QuickBird 영상의 적용 결과는 제안된 알고리즘이 무감독 영상 분류를 위한 강력한 수단임을 보여준다.

• 한국주거학회논문집
• /
• 제19권2호
• /
• pp.43-54
• /
• 2008

The aim of the research is to find out the direction of unit-plans in multi-dwellings for the future society. Shinonome Canal Court where residents actually live now are the objects in this study, and the residential floor plan of unit-plans were analyzed to find out the typical types. The analysis was focused on the unit-plans of 5 blocks of Shinonome Canal Court. Space Syntax Theory was used as the analysis method. As the first stage of the analysis, justified graphs were made to find out the characters of unit-plans through the classification of the graphs. Contents of the analysis are as follows: Relationship between classified justified graphs and dimension according to node number. Relationship between classified justified graph patterns and unit-plans. Characters of unit-plans in each blocks. Shinonome Canal Court consists of mainly small scale unit-plans and 30unit-plans are classified. 1LDK, 2LDK, 1LDK+S, 1LDK+f are typical unit-plans which are mainly supplied in the complex. It is noted that the results of the analysis by node, justified graph pattern and dimension are the same. It also presents diverse unit-plans which shows a change in nLDK pattern or add f (foyer), AN (annex), S (service room), Fs (free space) to basic nLDK type. In summary, it demonstrates the possibility of creating new residental floor plans in multi-dwellings.

• 한국인터넷방송통신학회논문지
• /
• 제23권6호
• /
• pp.89-95
• /
• 2023

인간 행동 인식 (Human action recognition, HAR) 기술은 스포츠 분석, 인간과 로봇 간의 상호작용, 대형 사이니지 콘텐츠 등의 애플리케이션에 활용되는 핵심 기술 중 하나이다. 본 논문에서는 몰입형 대형 사이니지 콘텐츠를 위한 STAGCN (Spatial temporal attention graph convolutional network) 기반 인간 행동 인식 시스템을 제안한다. STAGCN은 attention mechanism을 통해 스켈레톤 시퀀스의 시공간적 특징에 서로 다른 가중치를 부과하여, 동작 인식에 중요한 관절 및 시점을 고려할 수 있다. NTU RGB+D 데이터셋을 사용한 실험 결과, 제안된 시스템은 기존 딥러닝 모델들에 비해 높은 분류 정확도를 달성한 것을 확인했다.

• Korean Journal of Mathematics
• /
• 제20권2호
• /
• pp.247-254
• /
• 2012

The Klee-Quaife problem is finding the minimum order ${\mu}(d,c,v)$ of the $(d,c,v)$ graph, which is a $c$-vertex connected $v$-regular graph with diameter $d$. Many authors contributed finding ${\mu}(d,c,v)$ and they also enumerated and classied the graphs in several cases. This problem is naturally extended to the case of digraphs. So we are interested in the extended Klee-Quaife problem. In this paper, we deal with an equivalent problem, finding the maximum diameter of digraphs with given order, focused on 2-regular case. We show that the maximum diameter of strongly connected 2-regular digraphs with order $n$ is $n-3$, and classify the digraphs which have diameter $n-3$. All 15 nonisomorphic extremal digraphs are listed.

• 한국지능시스템학회논문지
• /
• 제16권2호
• /
• pp.228-233
• /
• 2006

This paper introduces design method for air pump system using bond graph and genetic programming to maximize outflow subject to a constraint specifying maximum power consumption. The air pump system is a mixed domain system which includes electromagnetic, mechanical and pneumatic elements. Therefore an appropriate approach for a better system for synthesis is required. Bond graphs are domain independent, allow free composition, and are efficient for classification and analysis of models. Genetic programming is well recognized as a powerful tool for open-ended search. The combination of these two powerful methods, BG/GP, was tested for redesign of air pump system.

• 지능정보연구
• /
• 제2권1호
• /
• pp.45-58
• /
• 1996

This paper describes an a, pp.oach to recognizing composite features of prismatic parts. AAG (Attribute Adjacency Graph) is adopted as the basis of describing basic feature, but it is extended to enhance the expressive power of AAG by adding face type, angles between faces and normal vectors. Our a, pp.oach is called Extended AAG (EAAG). To simplify the recognition procedure, feature classification tree is built using the graph types of EEA and the number of EAD's. Algorithms to find open faces and dimensions of features are exemplified and used in decomposing composite feature. The processing sequence of recognized features is automatically determined during the decomposition process of composite features.

• 정보처리학회논문지B
• /
• 제10B권3호
• /
• pp.265-272
• /
• 2003

본 논문에서는 동적으로 변화하는 대용량의 데이터베이스로부터 보다 현실적인 데이터 마이닝의 수행을 가능케 하기 위하여 기존의 직사각형분해 알고리즘을 개선한 새로운 알고리즘을 제안한다. 새로운 알고리즘은 이진 행렬을 이분(bipartite) 그래프로 변환하고, 변환된 이분 그래프에서 이분클리크(biclique)를 찾음으로써 직사각형 분해를 수행한다 제안된 알고리즘은 새롭게 유도된 수학적 정리들을 바탕으로 출발하였으며, 복잡도 분석을 통하여 그 효율성을 보이고, 기존의 분류 방법론과의 비교를 통하여 제안된 방법론이 규칙의 수와 분류율면에서 우수함을 보인다.

• 한국정보과학회 언어공학연구회:학술대회논문집(한글 및 한국어 정보처리)
• /
• 한국정보과학회언어공학연구회 2021년도 제33회 한글 및 한국어 정보처리 학술대회
• /
• pp.564-567
• /
• 2021

스미싱은 SMS 문자를 통해 피해자를 현혹시켜 개인정보나 금전 등을 갈취하는 범죄이다. 발전하는 스미싱 범죄 수법에 대응하기 위해선 새로운 스미싱 범죄 사례에서 데이터를 추출하고, 추출한 데이터를 기존 시스템에 통합하여 빠르게 대응할 수 있어야 한다. 본 연구에서는 빠른 스미싱 대응을 위해 전처리를 하지 않은 SMS 문자 텍스트에서 지식베이스를 자동으로 추출하고 저장하는 자동 지식베이스 추출 모듈을 제안하며, 추출 시스템 지식베이스를 바탕으로 입력된 SMS가 스미싱인지 판별하는 스미싱 SMS 탐지 모듈을 통합한 자동 지식베이스 추출 기반 스미싱 SMS 탐지 시스템을 제시한다. 제시된 스미싱 SMS 탐지 모델은 UCI SMS Spam Collection Dataset을 기준으로 90.9 (F1 score)의 성능을 보여주었다.