• 응용통계연구
• /
• 제15권2호
• /
• pp.251-267
• /
• 2002

Kohonen이 제안한 자기조직화 지도(Self Organizing Maps : SOM)는 매우 빠른 신경망 모형이다. 하지만 다른 신경망 모형과 마찬가지로 학습 결과에 대한 명확한 규칙을 제시하지 못할 뿐만 아니라 지역적 최적값으로 빠지는 경우가 종종 있다. 본 논문에서는 이러한 자기조직화 지도의 모형에 대한 설명력을 부여하고 전역 최적값으로 수렴할 수 있는 예측 성능을 갖는 모형으로서 자율학습 신경망에 베이지안 추론을 결합한 자기조직화 지도를 위한 베이지안 학습(Bayesian Learning for Self Organizing Maps ; BLSOM)을 제안한다. 이 방법은 기존의 자기조직화 지도가 지역적 해에 머물러 있는 것에 비해서 언제든지 전역적 해로 수렴함이 실험을 통하여 밝혀졌다.

• 한국정보시스템학회:학술대회논문집
• /
• 한국정보시스템학회 2000년도 추계학술대회
• /
• pp.25-35
• /
• 2000

A huge amount of financial information in large databases makes performance comparisons among organizations difficult or at least very time-consuming. This paper investigates whether neural networks in the form of self-organizing maps can be effectively employed to perform a competitive benchmarking in large databases. By using self-organizing maps, we expect to overcome problems associated with finding appropriate underlying distributions and functional forms of underlying data. The method also offers a way of visualizing the results. The database in this study consists of annual financial reports of 100 biggest Korean companies over the years 1998, 1999, and 2000.

• 전기전자학회논문지
• /
• 제12권4호
• /
• pp.217-224
• /
• 2008

This paper provides a framework for semantic correspondence of heterogeneous databases using self- organizing map. It solves the problem of overlapping between different databases due to their different schemas. Clustering technique using self-organizing maps (SOM) is tested and evaluated to assess its performance when using different kinds of data. Preprocessing of database is performed prior to clustering using edit distance algorithm, principal component analysis (PCA), and normalization function to identify the features necessary for clustering.

• 한국지능정보시스템학회:학술대회논문집
• /
• 한국지능정보시스템학회 1999년도 추계학술대회-지능형 정보기술과 미래조직 Information Technology and Future Organization
• /
• pp.303-311
• /
• 1999

The amount of financial information in today's sophisticated large data bases is huge and makes comparisons between company performance difficult or at least very time consuming. The purpose of this paper is to investigate whether neural networks in the form of self-organizing maps can be used to manage the complexity in large data bases. This paper structures and analyzes accounting numbers in a large data base over several time periods. By using self-organizing maps, we overcome the problems associated with finding the appropriate underlying distribution and the functional form of the underlying data in the structuring task that is often encountered, for example, when using cluster analysis. The method chosen also offers a way of visualizing the results. The data base in this study consists of annual reports of more than 80 Korean companies with data from the year 1998.

• Journal of applied mathematics & informatics
• /
• 제9권2호
• /
• pp.793-806
• /
• 2002

Feature-based similarity retrieval become an important research issue in image database systems. The features of image data are useful to discrimination of images. In this paper, we propose the highspeed k-Nearest Neighbor search algorithm based on Self-Organizing Maps. Self-Organizing Maps (SOM) provides a mapping from high dimensional feature vectors onto a two-dimensional space. The mapping preserves the topology of the feature vectors. The map is called topological feature map. A topological feature map preserves the mutual relations (similarity) in feature spaces of input data. and clusters mutually similar feature vectors in a neighboring nodes. Each node of the topological feature map holds a node vector and similar images that is closest to each node vector. In topological feature map, there are empty nodes in which no image is classified. We experiment on the performance of our algorithm using color feature vectors extracted from images. Promising results have been obtained in experiments.

• 한국경영과학회지
• /
• 제31권3호
• /
• pp.41-51
• /
• 2006

Clustering is to group similar objects into clusters. Until now there are a lot of approaches using Self-Organizing feature Maps (SOFMS) But they have problems with a small output-layer nodes and initial weight. For example, one of them is a one-dimension map of c output-layer nodes, if they want to make c clusters. This approach has problems to classify elaboratively. This Paper suggests one-dimensional output-layer nodes in SOFMs. The number of output-layer nodes is more than those of clusters intended to find and the order of output-layer nodes is ascending in the sum of the output-layer node's weight. We un find input data in SOFMs output node and classify input data in output nodes using Euclidean distance. The proposed algorithm was tested on well-known IRIS data and TSPLIB. The results of this computational study demonstrate the superiority of the proposed algorithm.

• 한국건축시공학회지
• /
• 제12권6호
• /
• pp.664-673
• /
• 2012

Determining the layout of temporary facilities that support construction activities at a site is an important planning activity, as layout can significantly affect cost, quality of work, safety, and other aspects of the project. The construction site layout problem involves difficult combinatorial optimization. Recently, various artificial intelligence(AI)-based algorithms have been applied to solving many complex optimization problems, including neural networks(NN), genetic algorithms(GA), and swarm intelligence(SI) which relates to the collective behavior of social systems such as honey bees and birds. This study proposes a site facility layout optimization algorithm based on self-organizing maps(SOM). Computational experiments are carried out to justify the efficiency of the proposed method and compare it with particle swarm optimization(PSO). The results show that the proposed algorithm can be efficiently employed to solve the problem of site layout.

• 한국경영과학회지
• /
• 제26권3호
• /
• pp.1-20
• /
• 2001

The amount of financial information in sophisticated large data bases is huge and makes interfirm performance comparisons very difficult or at least very time consuming. The purpose of this paper is to investigate whether neural networks in the form of self-organizing maps (SOM) can be successfully employed to manage the complexity for competitive financial benchmarking. SOM is known to be very effective to visualize results by projecting multi-dimensional financial data into two-dimensional output space. Using the SOM, we overcome the problems of finding an appropriate underlying distribution and the functional form of data when structuring and analyzing a large data base, and show an efficient procedure of competitive financial benchmarking through clustering firms on two-dimensional visual space according to their respective financial competitiveness. For the empirical purpose, we analyze the data base of annual reports of 100 Korean listed companies over the years 1998, 1999, and 2000.

• 대한토목학회논문집
• /
• 제29권3D호
• /
• pp.347-356
• /
• 2009

본 연구는 기존의 도로기능분류 정의와 방법론을 벗어나 교통특성에 따른 도로분류 방법론인 도로특성분류를 기초로 분석을 수행하였다. 도로특성분류에 대한 일련의 과정 중에서 다양한 교통특성을 반영하는 설명변수를 기초로 요인점수를 산출하고, 동질한 도로구간을 그룹핑하는 군집화 분석과정과 적정 군집수 도출에 따른 군집결과비교에 본 연구는 초점을 맞추었다. 도로분류를 위해 병합적 계층 군집분석인 Ward법, 비계층적 군집분석인 K-means법, 자율신경 회로망을 이용한 K-SOM을 사용하여 비교분석하였다. 각 군집기법에 대한 결과를 토대로 비교분석한 결과, 군집 수 5 이하에서는 K-means법, 군집 수 14 이상에서는 Kohonen selforganizing maps가 가장 우수한 것으로 나타났으며, 군집수 5~9사이에서는 Ward법과 Kmeans법의 군집 성능이 불규칙한 패턴을 보임에 따라 세밀한 결과분석을 통해 우수성을 결정하는 것이 바람직할 것으로 분석되었다. 본 연구결과는 다양한 교통특성을 고려한 도로구간의 군집 속성을 분석하고 예측하는 분류화 작업에 중요한 기초적인 자료로 사용될 것으로 기대된다.

• 한국정보과학회:학술대회논문집
• /
• 한국정보과학회 1998년도 가을 학술발표논문집 Vol.25 No.2 (2)
• /
• pp.183-184
• /
• 1998

본 논문에서는 목적어-동사 관계의 분포에 따라 한국어 동사를 자동적으로 클러스터링하는 방법을 제시한다. SOM(Self-Organizing Map)이 입력 패턴을 분석하고 가시화하는데 뛰어난 성능을 보이므로, 본 논문에서는 클러스터링하는 방법으로 SOM을 채택하였다. 일단 맵(map)이 만들어지고 나면 학습하는 동안 경험하지 못한 동사도 쉽게 적당한 클러스터로 분류될 수 있고 클러스터들 간의 의미 거리도 맵을 이용하여 쉽게 계산할 수 있다. 본 논문에서 제안한 방법을 명사 확률 분포의 상대 엔트로피(relative entropy)에 기반한 클러스터링 방법과 비교해 본 결과, SOM에 의해 만들어진 동사 클러스터가 상대 엔트로피를 이용해서 만들어진 클러스터를 잘 반영한다는 것을 알 수 있었다.