• Journal of the Korean Society of Clothing and Textiles
• /
• v.20 no.1
• /
• pp.170-182
• /
• 1996

The purpose of this study is to classify the upper body of women into several kind.; of somatotypes, using the method of Surgical Tape and making their shells. The subjects are 50 females 20 to 29 years-old. Fifty-one anthropometric data are measured per shell of body surface : eight somatotype factors are obtained through principal component analysis and orthogonal rotation by the method of Varimax, Somatotype of women's upper body is achieved by cluster analysis, using the standardized factor score a.: an independent variable and the FASTCLUS of SAS by Kmeans. The results are as follows : 1. The number of the factors which explain the somatotype is eight and these factors comprise 81.63 percent of total variance. Factor 1 related to the degree of fatness in the front of upper body Factor 2 related to the degree of fatness in the back of upper body Factor 3 . related to the length of the upper body Factor 4 : related to the type of the upper chest over the chest circmference line Factor 5 : related to the armhole and neck Factor 6 : related to the type of lower chest under the chest circumference line Factor 7: related to the part of the back shoulder Factor 8: related to the depth of front neck and side dart of front independently 2. Cluster analysis results in classification of upper body into five clusters. Cluster 1 : the of circumference i.: lager and that of length is longer than the average The louver part of chest is the lagest and widest among surface areas. Cluster 2 : the circumference is the smallest , the length and surface area are small. The upper and lower chest is small Cluster 3 : the length and surface area are the smallest , the circumference is average. The body line (silhouette) from chest to waist is curved slightly.

• Bulletin of the Korean Chemical Society
• /
• v.34 no.11
• /
• pp.3327-3334
• /
• 2013

Infrared multiphoton dissociation (IRMPD) spectra of various protonated 1,2-diaminoethane-water clusters DAE-$H^+-(H_2O)_n$ (n = 1-6) were measured in the wavelength range of 3000-3800 $cm^{-1}$. The IRMPD spectra of the well separated ionic clusters were simulated by the MP2 method employing various basis sets. Comparison of the IRMPD spectra with the theory indicates that each cluster may exist as several low-lying conformers, and the sum spectra of the various conformers reveal almost one to one correspondence between theory and experiment. Free N-H and O-H stretches are observed in the ranges of 3400-3500 and 3600-3800 $cm^{-1}$, respectively. The $O-H{\cdots}N$ and $N-H{\cdots}O$ stretches are, however, observed in the broad region of 3000-3600 $cm^{-1}$. The theoretical calculations on DAE-$H^+-(H_2O)_n$ (n = 1-4) show gradual decrease of the average binding energy between DAE-$H^+$ and $H_2O$ as the cluster size increases, attaining the lowest value of 55 kJ/mol when n = 4. We found a low energy barrier of 21 kJ/mol to the isomerization converting the lowest energy cluster of DAE-$H^+-(H_2O)_n$ to the second lowest one.

• Journal of the Korea Society of Computer and Information
• /
• v.19 no.7
• /
• pp.1-8
• /
• 2014

Clustering is one of the most popular unsupervised learning methods, which is widely used to form clusters with homogeneous data. Clustering was used to extract contexts corresponding to clusters and a classification method was applied to each context or cluster individually. However, it is difficult to say that the unsupervised clustering is the best context forming method from the view of classification. In this paper, a new clustering method considering classification was proposed. The proposed method tries to minimize classification error in each cluster when a classification method is applied to each context locally. For this purpose, the proposed method adds constraints forcing two data points belong to the same class to have small distances, and two data points belong to different classes to have large distances in each cluster like in linear discriminant analysis. The usefulness of the proposed method is confirmed by experimental results.

• KIPS Transactions on Computer and Communication Systems
• /
• v.4 no.11
• /
• pp.369-382
• /
• 2015

Energy aware server clusters aim to reduce power consumption at maximum while keeping QoS(quality of service) compared to energy non-aware server clusters. They adjust the power mode of each server in a fixed or variable time interval to activate only the minimum number of servers needed to handle current user requests. Previous studies on energy aware server cluster put efforts to reduce power consumption or heat dissipation, but they do not consider energy efficiency well. In this paper, we propose an energy efficient cluster management method to improve not only performance per watt but also QoS of the existing server power mode control method based on autonomous learning. Our proposed method is to adjust server power mode based on a hybrid approach of autonomous learning method with multi level thresholds and power consumption prediction method. Autonomous learning method with multi level thresholds is applied under normal load situation whereas power consumption prediction method is applied under abnormal load situation. The decision on whether current load is normal or abnormal depends on the ratio of the number of current user requests over the average number of user requests during recent past few minutes. Also, a dynamic shutdown method is additionally applied to shorten the time delay to make servers off. We performed experiments with a cluster of 16 servers using three different kinds of load patterns. The multi-threshold based learning method with prediction and dynamic shutdown shows the best result in terms of normalized QoS and performance per watt (valid responses). For banking load pattern, real load pattern, and virtual load pattern, the numbers of good response per watt in the proposed method increase by 1.66%, 2.9% and 3.84%, respectively, whereas QoS in the proposed method increase by 0.45%, 1.33% and 8.82%, respectively, compared to those in the existing autonomous learning method with single level threshold.

• Korean Business Review
• /
• v.18 no.2
• /
• pp.197-223
• /
• 2005

The industrial cluster policy of 21st century rise to the focus method of regional economic promotion, therefore, the importance of study in cluster identification and mapping as policy task will bring into relief. This paper will confirms the estimate index and policy of industrial clusters with regional industry. The result in this case study, Cluster development should embrace the pursuit of competitive advantage and specialization rather than simply imitate successful clusters in other locations. This requires building on local sources of uniqueness. Government should reinforce and building on existing and emerging clusters rather than attempt to create entirely new ones. This sort of role for government is very different from industrial policy. The aim of cluster policy is to reinforce the development of all clusters. Not all clusters will succeed, but market forces should determine the outcomes. In other words, government should build on market- oriented system and innovative infra. The result of this study is meaning to the development of objectivity estimate index and derivation of cluster-focused policy with a case study of industrial clusters.

• Convergence Security Journal
• /
• v.7 no.2
• /
• pp.107-118
• /
• 2007

Distributed wireless sensor network in various environment have characteristic that is surveillance of environment-element and offering usefully military information but there is shortcoming that have some secure risks. Therefore secure service must be required for this sensor network safety. More safe and effective techniques of node administration are required for safe communication between each node. This paper proposes effective cluster-header and clustering techniques in suitable administration techniques of group-key on sensor network. In this paper, first each node transmit residual electric power and authentication message to BS (Base-Station). BS reflects "Validity Authentication Rate" and residual electric power. And it selects node that is more than these regularity values by cluster header. After BS broadcasts information about cluster header in safety and it transmits making a list of information about cluster member node to cluster header. Also, Every rounds it reflects and accumulates "Validity Authentication Rate" of former round. Finally, BS can select more secure cluster header.

• Journal of the Korea Society of Computer and Information
• /
• v.13 no.4
• /
• pp.127-132
• /
• 2008

Wireless sensor network technologies applicable to various industrial fields are rapidly growing. Because it is difficult to change a battery for the once distributed wireless sensor network, energy efficient design is very critical. In order to achieve this purpose in network design, a number of studies have been examining the energy efficient routing protocol. The sensor network consumes energy in proportion to the distance of data transmission and the data to send. Cluster-based routing Protocols such as LEACH-C achieve energy efficiency through minimizing the distance of data transmission. In LEACH-C, however, the total distance between the nodes consisting the clusters are considered important in constructing clustering. This paper examines the cluster-head-selection-algorithm that reflect the distance between the base station and the cluster-head having a big influence on energy consumption. The Proposed method in this paper brought the result that the performance improved average $4{\sim}7%$ when LEACH-C and the base station are located beyond a certain distance. This result showed that the distance between cluster-head and the base station had a substantial influence on lifetime performance in the cluster-based routing protocol.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2016.05a
• /
• pp.645-648
• /
• 2016

Cluster based wireless sensor networks have a characteristic that the cluster heads collect and aggregate data from sensor nodes and send data to sink node. In addition, between the adjacent sensor nodes deployed in the same area is characterized to the similar sensing data. In this paper, we propose a transmission algorithm for improving the energy efficiency using these two features in the cluster-based wireless sensor networks. Adjacent neighboring nodes form a pair and the two nodes sense data on shifts for one round. Additionally, two cluster heads are selected in a cluster and one of them alternately collects data from nodes and transmits data to the sink. This paper describes a transmission rounding method and a transmission frame for increasing energy efficiency and compared with conventional methods. We perform computer simulations to evaluate the performance of the proposed algorithm, and show better performance in terms of energy efficiency as compared with the LEACH algorithm.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2014.10a
• /
• pp.620-623
• /
• 2014

Network topology for communication between vehicles are quickly changing because vehicles have a special movement pattern, especially character which is quickly changed by velocity and situation of road. Because of these feature, it is not easy to apply reliable routing on VANET(Vehicular Ad-hoc Network). Clustering method is one of the alternatives which are suggested for overcoming weakness of routing algorithm. Clustering is the way to communicate and manage vehicles by binding them around cluster head. Therefore choosing certain cluster head among vehicles has a decisive effect on decreasing overhead in relevant clustering and determining stability and efficiency of the network. This paper introduces new cluster head selection algorithm using RSU(Road Side Unit) different from existing algorithms. We suggest a more stable and efficient algorithm which decides a priority of cluster head by calculating vehicles' velocity and distance through RSU than existing algorithms.

• Proceedings of the Korean Society of Surveying, Geodesy, Photogrammetry, and Cartography Conference
• /
• 2007.04a
• /
• pp.121-125
• /
• 2007

Recently, constructing and complementing the road network database are a main key in all social operation in our life. However it needs high expenses for constructing and complementing the data, and relies on many people for finishing the tasks. This study propose a novel method to extract urban road networks from 3-D LiDAR data automatically. This method integrates height, reflectance, and clustered road point information. Geometric information of general roads is also applied to cluster road points group correctly. The proposed method has been tested on various urban areas which contain complicated road networks. The results conclude that the integration of height, reflectance, and geometric information worked reliably to cluster road points.