• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 2006.05a
• /
• pp.305-312
• /
• 2006

K-means clustering algorithm 에서 주로 이루어지는 랜덤 초기화 (random initialization) 방법은 전역 최적화된 해(global minimum)를 찾아내기에 문제점을 지니고 있다. 즉, 여러 횟수의 알고리듬 반복(iteration)을 실행하더라도 전역 최적화된 해를 찾아내기가 매우 힘들며 주어진 자료의 크기(data size)가 큰 경우에 있어서 이는 거의 불가능하다. 본 논문은 이러한 문제점들을 극복하기 위한 방안으로, wavelet을 이용하여 최적의 초기 군집 중심점(initial clustering center)들을 선택하는 방법을 제시한다. 즉, 웨이블릿을 이용한 효과적인 초기화 (initialization)를 통해서 작은 알고리듬 반복 횟수만으로도 전역 최적화에 도달하는 초기화 방법을 기술한다. 이런 초기화 방법이 군집 알고리즘에 사용될 경우, 온라인상에서 실시간 이루어지는 군집 분석에 큰 도움이 된 수 있다.

• Journal of Digital Convergence
• /
• v.12 no.10
• /
• pp.243-252
• /
• 2014

Many researches about the effect on politics, economics and Sociocultural phenomenon using the social media are in progress. Authors utilized NAVER Trend most famous web browsing service in korea, NAVER Blog social media, NAVER Cafe service and Open Data(API) and also used temperature, humidity index data of Korea Meteorological Administration. This study analyzed a change of the public's emotion in korea using Cluster analysis of vocabulary of taste among its of feelings and senses. K-means clustering was followed by decision of the number of groups which was used Chi-square goodness of fit test and ward analysis. Eight groups was made and it represented sensitive vocabulary. By Discriminant analysis, eight groups decided by Cluster analysis has 98.9% accuracy. The change of the public's emotion has capability to predict people's activity so they can share sensibility and a bond of sympathy developed between them.

• Journal of Korea Society of Digital Industry and Information Management
• /
• v.19 no.4
• /
• pp.1-11
• /
• 2023

Clustering analysis is used in various fields including customer segmentation and clustering methods such as k-means are actively applied in the credit card customer segmentation. In this paper, we summarized the input features selection method of k-means clustering for the case of the credit card customer segmentation problem, and evaluated its feasibility through the analysis results. By using the label values of k-means clustering results as target features of a decision tree classification, we composed a method for prioritizing input features using the information gain of the branch. It is not easy to determine effectiveness with the clustering effectiveness index, but in the case of the CH index, cluster effectiveness is improved evidently in the method presented in this paper compared to the case of randomly determining priorities. The suggested method can be used for effectiveness of actively used clustering analysis including k-means method.

• Proceedings of the Korean Statistical Society Conference
• /
• 2005.11a
• /
• pp.97-100
• /
• 2005

A new model-based clustering algorithm is proposed. The idea starts from the assumption that observations are realizations of Gaussian processes and so are correlated. With a special covariance structure, the posterior probability that an observation belongs to each cluster is computed using the ECM algorithm. A preliminary result of small-scale simulation study is given to compare with the k-means clustering algorithms.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.13 no.1
• /
• pp.12-17
• /
• 2003

Optimal determination of cluster size has an effect on the result of clustering. In K-means algorithm, the difference of clustering performance is large by initial K. But the initial cluster size is determined by prior knowledge or subjectivity in most clustering process. This subjective determination may not be optimal. In this Paper, the genetic algorithm based optimal determination approach of cluster size is proposed for automatic determination of cluster size and performance upgrading of its result. The initial population based on attribution is generated for searching optimal cluster size. The fitness value is defined the inverse of dissimilarity summation. So this is converged to upgraded total performance. The mutation operation is used for local minima problem. Finally, the re-sampling of bootstrapping is used for computational time cost.

• Journal of the Korea Society of Computer and Information
• /
• v.29 no.5
• /
• pp.155-164
• /
• 2024

This study proposes an unsupervised learning-based clustering model to estimate the ESG ratings of domestic public institutions. To achieve this, the optimal number of clusters was determined by comparing spectral clustering and k-means clustering. These results are guaranteed by calculating the Davies-Bouldin Index (DBI), a model performance index. The DBI values were 0.734 for spectral clustering and 1.715 for k-means clustering, indicating lower values showed better performance. Thus, the superiority of spectral clustering was confirmed. Furthermore, T-test and ANOVA were used to reveal statistically significant differences between ESG non-financial data, and correlation coefficients were used to confirm the relationships between ESG indicators. Based on these results, this study suggests the possibility of estimating the ESG performance ranking of each public institution without existing ESG ratings. This is achieved by calculating the optimal number of clusters, and then determining the sum of averages of the ESG data within each cluster. Therefore, the proposed model can be employed to evaluate the ESG ratings of various domestic public institutions, and it is expected to be useful in domestic sustainable management practice and performance management.

• Journal of Auto-vehicle Safety Association
• /
• v.13 no.3
• /
• pp.69-77
• /
• 2021

Although accident data from the National Police Agency and insurance companies do not know the vehicle safety, the damage level information can be obtained from the data managed by the bus credit association or the bus company itself. So the accident severity was analyzed based on the side impact accidents using accident repair cost. K-means clustering analysis separated the cost of accident repair into 'minor', 'moderate', 'severe', and 'very severe'. In addition, the side impact accident severity was analyzed by using an ordered logit model. As a result, it is appeared that the longer the repair period, the greater the impact on the severity of the side impact accident. Also, it is appeared that the higher the number of collision points, the greater the impact on the severity of the side impact accident. In addition, oblique collisions of the angle of impact were derived to affect the severity of the accident less than right angle collisions. Finally, the absence of opponent vehicle and large commercial vehicles involved accidents were shown to have less impact on the side impact accident severity than passenger cars.

• Spatial Information Research
• /
• v.20 no.6
• /
• pp.13-21
• /
• 2012

In this study, homogeneous regions for regional frequency analysis were identified using rainfall data from 61 observation points in Korea. The used data were gathered from 1980 to 2010. Self organizing map and K-means clustering based on Davies-Bouldin Index were used to make clusters showing similar rainfall patterns and to decide the optimum number of the homogeneous regions. The results from this analysis showed that the 61 observation points can be optimally grouped into 6 geographical clusters. Finally, the 61 observations points grouped into 6 clusters were mapped regionally using Thiessen polygon method.

• Proceedings of the Korean Information Science Society Conference
• /
• 2011.06a
• /
• pp.33-36
• /
• 2011

클러스터링은 데이터 포인트들을 그룹으로 묶어 데이터를 분석하는데 유용하다. 특히 K-means는 가장 널리 쓰이는 클러스터링 알고리즘으로 k개의 군집(Cluster)을 찾는다. 본 논문에서는 기존의 K-means 알고리즘과 비교해 고차원 대규모데이터에 대해서 효율적으로 동작하는 K-means 알고리즘을 제안한다. 제안된 알고리즘은 기존의 알고리즘에서와 같이 거리 정보를 이용해 불필요한 계산을 줄여나가며 또한 움직임 없는 군집들을 계산에서 제외하여 수행시간을 단축한다. 제안된 알고리즘은 기존의 관련연구에서 제안된 알고리즘에 비해 공간을 적게 쓰면서 동시에 빠르다. 실제 고차원 데이터 실험을 통해서 제안된 알고리즘의 효율성을 보였다.

• The KIPS Transactions:PartD
• /
• v.12D no.3 s.99
• /
• pp.461-470
• /
• 2005

This paper suggests a new policy for consolidating a company's profits by segregating the clients using the contents service and allocating the media server's resources distinctively by clusters using the cluster analysis method of CRM, which is mainly applied to marketing. In this case, CRM refers to the strategy of consolidating a company's profits by efficiently managing the clients, providing them with a more effective, personalized service, and managing the resources more effectively. For the realization of a new service policy, this paper analyzes the level of contribution $vis-\acute{a}-vis$ the clients' service pattern (total number of visits to the homepage, service type, service usage period, total payment, average service period, service charge per homepage visit) and profits through the cluster analysis of clients' data applying the K-Means Method. Clients were grouped into 4 clusters according to the contribution level in terms of profits. Likewise, the CRFA (Client Request Filtering algorithm) was suggested per cluster to allocate media server resources. CRFA issues approval within the resource limit of the cluster where the client belongs. In addition, to evaluate the efficiency of CRFA within the Client/Server environment the acceptance rate per class was determined, and an evaluation experiment on network traffic was conducted before and after applying CRFA. The results of the experiments showed that the application of CRFA led to the decrease in network expenses and growth of the acceptance rate of clients belonging to the cluster as well as the significant increase in the profits of the company.