• 경영과학
• /
• 제19권1호
• /
• pp.189-196
• /
• 2002

Data clustering is often one of the first steps in data mining analysis. It Identifies groups of related objects that can be used as a starling point for exploring further relationships. This technique supports the development of population segmentation models, such as demographic-based customer segmentation. This paper Purpose to present the development of two phase hierarchical clustering algorithm for group formation. Applications of the algorithm for product-customer group formation in customer relationahip management are also discussed. As a result of computer simulations, suggested algorithm outperforms single link method and k-means clustering.

• 한국지능시스템학회논문지
• /
• 제22권6호
• /
• pp.729-734
• /
• 2012

패턴 인식, 데이터 분석, 침입 탐지, 이미지 처리, 바이오 인포매틱스 등과 같은 수많은 분야에서 클러스터링 분석이 사용되고 있다. 기존의 많은 연구들은 수치 데이터에만 기반을 두고 있다. 그러나 구간 데이터, 히스토그램, 심지어는 함수들을 값으로 갖는 변수들을 다루는 심볼릭 데이터 분석이 부상하고 있다. 본 논문에서는 이런 심볼릭 데이터들을 클러스터링하기 위하여 비대칭적 유사도를 제안한다. 또한 평균 유사도 값(ASV)에 기반한 클러스터링 방법도 개발한다. 제안하는 클러스터링의 결과는 기존 방법들과 다르며, 매우 고무적인 결과를 보여준다.

• Journal of the Korean Data and Information Science Society
• /
• 제17권2호
• /
• pp.301-309
• /
• 2006

In this paper, we show the improved fuzzy c-means clustering method. To improve, we use the double clustering as principal component analysis from objects which is located on common region of more than two clusters. In addition we use the degree of membership (probability) of fuzzy c-means which is the advantage. From simulation result, we find some improvement of accuracy in data of the probability 0.7 exterior and interior of overlapped area.

• 한국경영과학회지
• /
• 제31권2호
• /
• pp.87-98
• /
• 2006

In many cases of cluster analysis we are forced to perform clustering without any prior knowledge on the number of clusters. But in some clustering methods such as k-means algorithm it is required to provide the number of clusters beforehand. In this study, we focus on the problem to determine the number of clusters in the given data. We follow the 2 stage approach of ASA clustering algorithm and mainly try to improve the performance of the first stage of the algorithm. We verify the usefulness of the method by applying it for various kinds of simulated data. Also, we apply the method for clustering two kinds of real life qualitative data.

• 산업경영시스템학회지
• /
• 제37권4호
• /
• pp.239-244
• /
• 2014

The conventional clustering approaches are mostly based on minimizing total dissimilarity of input and output. However, the clustering approach may not be helpful in some cases of clustering decision making units (DMUs) with production feature converting multiple inputs into multiple outputs because it does not care converting functions. Data envelopment analysis (DEA) has been widely applied for efficiency estimation of such DMUs since it has non-parametric characteristics. We propose a new clustering method to identify groups of DMUs that are similar in terms of their input-output profiles. A real world example is given to explain the use and effectiveness of the proposed method. And we calculate similarity value between its result and the result of a conventional clustering method applied to the example. After the efficiency value was added to input of K-means algorithm, we calculate new similarity value and compare it with the previous one.

• Communications for Statistical Applications and Methods
• /
• 제27권6호
• /
• pp.589-602
• /
• 2020

The development of smart grids has enabled the easy collection of a large amount of power data. There are some common patterns that make it useful to cluster power consumption patterns when analyzing s power big data. In this paper, clustering analysis is based on distance functions for time series and clustering algorithms to discover patterns for power consumption data. In clustering, we use 10 distance measures to find the clusters that consider the characteristics of time series data. A simulation study is done to compare the distance measures for clustering. Cluster validity measures are also calculated and compared such as error rate, similarity index, Dunn index and silhouette values. Real power consumption data are used for clustering, with five distance measures whose performances are better than others in the simulation.

• 한국지하수토양환경학회:학술대회논문집
• /
• 한국지하수토양환경학회 2005년도 총회 및 춘계학술발표회
• /
• pp.296-299
• /
• 2005

Time series of spring discharge data in Jeju island can provide abundant information on the spatial groundwater system. In this study, the classification based on time series of spring discharge was performed with clustering analysis: discharge rate and EC. Peak discharges are mainly observed in august or september. However, double peaks and late peaks of discharge are also observed at a plenty of springs. Based on results of clustering analysis, it can be deduced that GH model is not appropriate for the conceptual model of Groundwater system in Jeju island. EC distributions in dry season are also support the conclusion.

• 한국문헌정보학회지
• /
• 제38권4호
• /
• pp.179-192
• /
• 2004

본 논문에서는 복합명사에 대한 색인 방법을 다각적으로 적용하여 계층적 결합 문서 클러스터링 시스템의 결과를 분석한다. 우선 한글 색인 엔진과 HAC(Hierarchical Agglomerative Clustering) 엔진에 대해서 설명하고 한글 색인 엔진에서 제공되는 3가지 복합명사 분석 모드에 대해서 기술한다. 또한 구현된 클러스터링 엔진의 특징과 속도 향상을 위한 기법 등을 예시한다. 실험에서는 3가지 복합명사 색인 방법을 기준으로 문서 클러스터링을 수행하고, 실험 결과에 대한 분석에서 복합명사에 대한 색인 방법이 문서 클러스터링의 결과에 직접적인 영향을 준다는 것을 보여준다.

• 정보관리학회지
• /
• 제23권4호
• /
• pp.215-231
• /
• 2006

패스파인더 네트워크를 사용하여 지적 구조의 분석과 규명을 시도한 여러 연구가 발표되었다. 패스파인더 네트워크는 다차원척도법에 비해서 여러 장점을 가지고 있지만 구축 알고리즘의 복잡도가 매우 높아서 실행 시간이 오래 걸리며, 전통적인 지적 구조 분석에 유용하게 사용되어온 군집분석을 함께 적용하기가 어려운 것이 단점이다. 이 연구에서는 이와 같은 패스파인더 네트워크의 약점을 보완할 수 있는 새로운 기법으로 병렬최근접이웃클러스터링(PNNC) 기법을 제안하였다. PNNC 기법의 클러스터링 성능을 전통적인 계층적 병합식 클러스터링 기법들과 비교해본 결과 효과성과 효율성 양면에서 기존 기법보다 우세한 것으로 확인되었다.

• 지식경영연구
• /
• 제17권3호
• /
• pp.65-91
• /
• 2016

Although the success of industrial clusters largely depends on whether clustering firms can achieve economic performance, there has been less attention on investigating factors and conditions contributing to the performance enhancement for clustering small and medium-sized enterprises (SMEs). Along this vein, we adopt the theories of social capital and organizational learning as those success factors for clustering SMEs. This study thus aims at examining what effect social capital accrued in the relationships among actors within clusters has on firm performance of clustering SMEs and what role organizational learning plays in the linkage between social capital and firm performance. For the empirical analysis, we operationalized the variables and their measures to develop questionnaires through the theoretical reviews on literatures. As a sample of 227 clustering SMEs, our collected data was analyzed by hierarchical regression analysis. The results confirmed that a high level of social capital, represented by network, trust, and norm, has positive effect on firm performance of clustering SMEs. We also found that clustering firms presenting high organizational learning, represented by absorptive and transformative capability, achieve better performance than those placing less value on organizational learning. Furthermore the significant relationship between social capital and firm performance is mediated partially through organizational learning. These findings imply not only that the territorial agglomeration of industrial cluster does not guarantee the performance creation of clustering SMEs but that they need to develop social capital among various actors within clusters, facilitating their knowledge diffusion. In order to absorb and mobilize the shared knowledge and information into strategic resources, the firms should improve their capability associated with organizational learning. These expand our understanding on the importance of social capital and organizational learning for the performance enhancement of clustering firms. Differentiating from major studies addressing benefits and advantages of industrial cluster, this study based on the perspective of firm-internal business process contributes to the literature advancement. Strategic and policy implications of this study are discussed in detail.