• Journal of Communications and Networks
• /
• v.8 no.3
• /
• pp.330-338
• /
• 2006

Many web cache systems and policies concerning them have been proposed. These studies, however, consider large objects less useful than small objects in terms of performance, and evict them as soon as possible. Even if this approach increases the hit rate, the byte hit rate decreases and the connections occurring over congested links to outside networks waste more bandwidth in obtaining large objects. This paper puts forth a client-cluster approach for improving the web cache system. The client-cluster is composed of the residual resources of clients and utilizes them as exclusive storage for large objects. This proposed system achieves not only a high hit rate but also a high byte hit rate, while reducing outgoing traffic. The distributed hash table (DHT) based peer-to-peer lookup protocol is utilized to manage the client-cluster. With the natural characteristics of this protocol, the proposed system with the client-cluster is self-organizing, fault-tolerant, well-balanced, and scalable. Additionally, the large objects are managed with an index based allocation method, which balances the loads of all clients well. The performance of the cache system is examined via a trace driven simulation and an effective enhancement of the proxy cache performance is demonstrated.

• Journal of Periodontal and Implant Science
• /
• v.51 no.1
• /
• pp.30-39
• /
• 2021

Purpose: The present study was undertaken to examine whether periodontal probe visibility (PV) accurately reflects gingival thickness (GT) and to identify factors affecting PV using cluster and multivariate analyses. Methods: The clinical characteristics of the maxillary central incisors (n=90 subjects) were examined. Clinical photographs, sex, PV, probing depth, gingival width, papilla height, GT as measured with an ultrasonic device, and the ratio of crown width to crown length were recorded. Multivariate analysis, using multinomial baseline-category logistic regression, was used to identify factors predictive of PV. Cluster analysis was used to identify gingival biotypes. Results: In the multivariate analysis, sex was the only significant predictor of PV (odds ratio, 6.48). Two clusters of subjects were created based on morphometric parameters. The mean GT among cluster A subjects was significantly lower than that among cluster B subjects (P=0.015). No significant difference was found between cluster A and B subjects in terms of PV score (P=0.583). Conclusions: Periodontal PV was not associated with GT as measured directly using an ultrasonic device. Sex was a highly significant predictor of periodontal PV. GT was found to be correlated with morphological characteristics of the periodontium.

• Journal of the Korean Operations Research and Management Science Society
• /
• v.42 no.3
• /
• pp.25-34
• /
• 2017

K-means is a popular and efficient data clustering method that only uses intra-cluster distance to establish a valid index with a previously fixed number of clusters. K-means is useless without a suitable number of clusters for unsupervised data. This paper aimsto propose the Group Search Optimization (GSO) using Silhouette to find the optimal data clustering solution with a number of clusters for unsupervised data. Silhouette can be used as valid index to decide the number of clusters and optimal solution by simultaneously considering intra- and inter-cluster distances. The performance of GSO using Silhouette is validated through several experiment and analysis of data sets.

• World Technopolis Review
• /
• v.6 no.2
• /
• pp.130-147
• /
• 2017

This paper aims to identify the spatial concentrations and linkage properties of industrial clusters in the Chungbuk Province region in Korea using a three-step approach, which is composed of the cluster index, Getis-Ord's $Gi^*$, and qualitative input-output analysis. The results of the study reveal: a) what industrial sectors are concentrated and where they are; b) where the spatially interdependent industries are; and c) how the industrial sectors of the identified clusters in different locations are vertically interconnected. In addition, the degree of strength of the interindustry linkages between industrial clusters are assessed. Based on the findings, some plausible industrial policies are suggested.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2005.11a
• /
• pp.171-174
• /
• 2005

기존의 클러스터 평가 지표(cluster validation index)는 클러스터의 개수가 커질수록 클러스터 평가 지표 값이 단조 감소하는 경향을 보인다. 최근에 이러한 단점을 보완하는 새로운 클러스터 평가 지표가 본 논문 저자중의 하나에 의해 제안되었으나, over-clustering의 단점 을 지니고 있다. 본 논문에서는, 클러스터 평가 지표 값이 단조 감소 및 over-clustering을 방지할 수 있는 새로운 클러스터 평가 지표를 제안하고, 여러 가지 예제를 통하여 새롭게 제안된 평가 지표의 타당성을 보인다.

• The Korean Journal of Ecology
• /
• v.5 no.2_3
• /
• pp.132-135
• /
• 1982

Structure and dynamics of the phytoplankton communities of Uiam lake, Korea, was investigated. In the Uiam lake four dominant species were found Oscillatoria limosa at Chuncheon City side, O. tenuis at Soyang river side, Melosira italica at south-east side and Asterionella gracillima at west side of the lake. By cluster analysis, based on the similarity index and dissimilarity index, the phytoplanktons in this lake were grouped into three communities; i.e. Oscillatoria, Melosira and Asterionella community. And also the same groups obtained by the cluster analysis were recognized by polar ordination technique along polluted degree gradient. It is clear that oscillatoria community occur in polluted site, Asterionella community in unpolluted site and Melosira community in less polluted site.

• Communications for Statistical Applications and Methods
• /
• v.22 no.3
• /
• pp.265-276
• /
• 2015

In this paper, we propose hierarchical cluster analysis and multidimensional scaling for joint distribution valued data. Information technology is increasing the necessity of statistical methods for large and complex data. Symbolic Data Analysis (SDA) is an attractive framework for the data. In SDA, target objects are typically represented by aggregated data. Most methods on SDA deal with objects represented as intervals and histograms. However, those methods cannot consider information among variables including correlation. In addition, objects represented as a joint distribution can contain information among variables. Therefore, we focus on methods for joint distribution valued data. We expanded the two well-known exploratory methods using the dissimilarities adopted Hall Type relative projection index among joint distribution valued data. We show a simulation study and an actual example of proposed methods.

• Journal of the Korean Society of Clothing and Textiles
• /
• v.19 no.5
• /
• pp.842-852
• /
• 1995

The purpose of this study was to provide fundamental data of characterizing for age group of obese women's body shape. The subjects were 132 obese women, age of 20~59. The criteria of obestity based on Vervaeck index ($\geq$92.3), Rohrer index ($\geq$ 1.6), and bust girth ($\geq$90 cm). Thirty eight measurement items were used for F-test. Forty two transformed variables (including 4 additional variables) were used for cluster analysis The results were as follows: 1. Characteristics for age group were described that 20's was narrow shoulder, flesh contoured shape in upper body. 30's was rounded upper back, 40's was the slightest fatty shape, and 50's was the fattest round shape among four age groups. 2. Four types of obese women's body shape were classified by cluster analysis. Body shape of type 1 was fatty abdomen, and rounded upper back in 50's mainly. Body shape of type 2 was slightly fatty. Body shape of type 3 was round shape. Body shape of type 4 was the smallest girth in the obese women in 40's mainly.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.10 no.3
• /
• pp.1091-1110
• /
• 2016

Hybrid deployment under current cloud data centers is a combination of online and offline services, which improves the utilization of the cluster resources. However, the performance of the cluster is often affected by the online services in the hybrid deployment environment. To improve the response time of online service (e.g. search engine), an effective scheduling algorithm based on Storm is proposed. At the component level, the algorithm dispatches the component with more influence to the optimal performance node. Inside the component, a reasonable resource allocation strategy is used. By searching the compressed index first and then filtering the complete index, the execution speed of the component is improved with similar accuracy. Experiments show that our algorithm can guarantee search accuracy of 95.94%, while increasing the response speed by 68.03%.

• The Korean Journal of Ecology
• /
• v.8 no.4
• /
• pp.205-214
• /
• 1985

At the Changnyeung natural bod, the flood in the freshwater wetland and the adaptation of the insect at this area were investigated by the species diversity index and the cluster analysis. Most dominant species was Diplonychus esakii collected at the site of the water edge and its dominance index was 0.797. This area had lower species diversity indices with the value of 0.340 to 1.712 than that of the grassland and water stream. The number of the species in this area was not rich. Some ground beetles inhabited at the flooded or wet area were important species for the pair group between the sites by the cluster analysis. The composition of the insect species at this natural wetland affected by the irregular flood was very simple and specific.