• The Journal of Society for e-Business Studies
• /
• v.22 no.3
• /
• pp.87-103
• /
• 2017

In recent years, many small and medium-sized manufacturing companies are making process innovation and product innovation through the public knowledge services. K Research institute provides different types of knowledge services in combination and due to this complexity, it is difficult to analyze the performance of knowledge service programs precisely. In this study, we derived performance items from bottom-up viewpoints, rather than top-down approaches selecting those items as in previous performance analysis. As a result, 74 items were finded from 82 companies in the K Research Institute case book, and the final result was refined to 17 items. After that a case-performance matrix was constructed, and binary data was entered to analyze. As a result, three clusters were identified through K-means clustering as 'enhancement of core competitiveness (product and patent),' 'expansion of domestic and overseas market,' and 'improvement of operational efficiency.'

• Fashion & Textile Research Journal
• /
• v.10 no.5
• /
• pp.644-651
• /
• 2008

The health of foot is connected with individual's health and affects men's activity. In order to develope comfort socks, both foot size and foot shape must be considered. The purpose of this study was to categorize men's foot shape according to age using men's foot scan data (with 2005 Size Korea). Factor analysis, Cluster analysis, ANOVA, and Duncan's test were performed for statistical analysis of the data by SPSS Win 12.00 program. The results are as follows. 1. Nine factors constituting the men's foot were extracted through factor analysis and those factors comprised 77.7% of total variance. 2. On the basis of the cluster analysis, four different foot shapes were categorized. Cluster 1 was characterized by large in toe and ankle size. Cluster 2 was characterized by short foot length, low foot height, and small foot breadth/girth. Cluster 3 was characterized by large and high in foot height. Cluster 4 was characterized by short in foot length and large in foot breadth/girth. 3. Distribution of four foot shape clusters from 20 to 70 years in age above were categorized. For the 20 to 29 years in age, cluster 2, while for the over 30 years in age cluster 4 or cluster 3 is the most dominant foot type. A foot breadth in the 50 years over is wider size range than that in the below 49 years. The foot figures of elderly men over 60 years were smaller than those of below 60 years.

• Journal of Navigation and Port Research
• /
• v.35 no.1
• /
• pp.57-62
• /
• 2011

On the view point of renewable energies as energy sources of district heating and cooling plant, the purpose of this study is to develop, classify and map the 500m${\times}$500m mesh, of which is treated as normal size in DHC regulations for evaluation process. Followings are the results. Various building and geographical informations including 13 districts and 108 counties are re-defined to create 500m${\times}$500m meshes, and it is find out that 3,289 meshes among 8,463 meshes have meaningful floor areas. Only 59 meshes(1.8%) are evaluated as mesh which has more than 50% of building volume ratio per mesh. 5 clusters classified by principal analysis and cluster analysis with building needs' characteristics are defined. Gwang-an Dong is representative of cluster 1 characterized as commercial area, and the cluster 4, 5 which has mainly residential needs are distributed in Yong-ho dong. Because there are a lot of cluster 3 meshes, which has complex needs area based on residential, cluster 3 could be defined as representative of Busan metropolitan city.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.12 no.1
• /
• pp.459-466
• /
• 2011

In this paper, we proposed a data mining framework for the management of alerts in order to improve the performance of the intrusion detection systems. The proposed alert data mining framework performs alert correlation analysis by using mining tasks such as axis-based association rule, axis-based frequent episodes and order-based clustering. It also provides the capability of classify false alarms in order to reduce false alarms. We also analyzed the characteristics of the proposed system through the implementation and evaluation of the proposed system. The proposed alert data mining framework performs not only the alert correlation analysis but also the false alarm classification. The alert data mining framework can find out the unknown patterns of the alerts. It also can be applied to predict attacks in progress and to understand logical steps and strategies behind series of attacks using sequences of clusters and to classify false alerts from intrusion detection system. The final rules that were generated by alert data mining framework can be used to the real time response of the intrusion detection system.

• The KIPS Transactions:PartB
• /
• v.8B no.5
• /
• pp.515-522
• /
• 2001

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 Map(SOM) provides a mapping from high dimensional feature vectors onto a two-dimensional space. 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. We implemented about k-NN search for similar image classification as to (1) access to topological feature map, and (2) apply to pruning strategy of high speed search. We experiment on the performance of our algorithm using color feature vectors extracted from images. Promising results have been obtained in experiments.

• Journal of the Korean Data and Information Science Society
• /
• v.20 no.3
• /
• pp.551-562
• /
• 2009

In this paper, I would present a using method for The Fishing Operation Information(FOI) of National Federation of Fisheries Cooperatives(NFFC) through the availabilities analysis and put out the similarities by the section of the sea through classifying characteristics of fishing patterns by their locations. As a result, although the catch of FOI is nothing more than 33% level to National Fishery Production Statistics(NFPS), FOI data is useful in understanding the patterns of fishing operation by the location because both patterns and correlation were very similar in the usability analysis, comparing the FOI data with NFPS. So I classified optimal clusters for catch, the number of fishing days and the number of fishing vessels through 2 step cluster analysis by the big marine zone and divided fishing patterns.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.26 no.5
• /
• pp.755-762
• /
• 2022

This paper propose a clustered algorithm that possible more efficient COVID-19 disease learning prediction within clustering using context-aware attribute information. In typically, clustering of COVID-19 diseases provides to classify interrelationships within disease cluster information in the clustering process. The clustering data will be as a degrade factor if new or newly processing information during treated as contaminated factors in comparative interrelationships information. In this paper, we have shown the solving the problems and developed a clustering algorithm that can extracting disease correlation information in using K-means algorithm. According to their attributes from disease clusters using accumulated information and interrelationships clustering, the proposed algorithm analyzes the disease correlation clustering possible and centering points. The proposed algorithm showed improved adaptability to prediction accuracy of the classification management system in terms of learning as a group of multiple disease attribute information of COVID-19 through the applied simulation results.

• Journal of Microbiology and Biotechnology
• /
• v.21 no.1
• /
• pp.5-13
• /
• 2011

In this study, seven Trichoderma species (33 strains) were classified using secondary metabolite profile-based chemotaxonomy. Secondary metabolites were analyzed by liquid chromatography-electrospray ionization tandem mass spectrometry (LC-ESI-MS-MS) and multivariate statistical methods. T. longibrachiatum and T. virens were independently clustered based on both internal transcribed spacer (ITS) sequence and secondary metabolite analyses. T. harzianum formed three subclusters in the ITS-based phylogenetic tree and two subclusters in the metabolitebased dendrogram. In contrast, T. koningii and T. atroviride strains were mixed in one cluster in the phylogenetic tree, whereas T. koningii was grouped in a different subcluster from T. atroviride and T. hamatum in the chemotaxonomic tree. Partial least-squares discriminant analysis (PLS-DA) was applied to determine which metabolites were responsible for the clustering patterns observed for the different Trichoderma strains. The metabolites were hetelidic acid, sorbicillinol, trichodermanone C, giocladic acid, bisorbicillinol, and three unidentified compounds in the comparison of T. virens and T. longibrachiatum; harzianic acid, demethylharzianic acid, homoharzianic acid, and three unidentified compounds in T. harzianum I and II; and koninginin B, E, and D, and six unidentified compounds in T. koningii and T. atroviride. The results of this study demonstrate that secondary metabolite profiling-based chemotaxonomy has distinct advantages relative to ITS-based classification, since it identified new Trichoderma clusters that were not found using the latter approach.

• Journal of Broadcast Engineering
• /
• v.18 no.4
• /
• pp.643-646
• /
• 2013

We present an improved AdaBoost algorithm to avoid overfitting phenomenon. AdaBoost is widely known as one of the best solutions for object detection. However, AdaBoost tends to be overfitting when a training dataset has noisy samples. To avoid the overfitting phenomenon of AdaBoost, the proposed method divides positive samples into K clusters using k-means algorithm, and then uses only one cluster to minimize the training error at each iteration of weak learning. Through this, excessive partitions of samples are prevented. Also, noisy samples are excluded for the training of weak learners so that the overfitting phenomenon is effectively reduced. In our experiment, the proposed method shows better classification and generalization ability than conventional boosting algorithms with various real world datasets.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.13 no.6
• /
• pp.661-666
• /
• 2003

In this paper, the conventional k-modes and fuzzy k-modes algorithms for clustering categorical data is extended by representing the clusters of categorical data with fuzzy centroids instead of the hard-type centroids used in the original algorithm. The hard-type centroids of the traditional algorithms had difficulties in dealing with ambiguous boundary data, which might be misclassified and lead to thelocal optima. Use of fuzzy centroids makes it possible to fully exploit the power of fuzzy sets in representing the uncertainty in the classification of categorical data. The distance measure between data and fuzzy centroids is more precise and effective than those of the k-modes and fuzzy k-modes. To test the proposed approach, the proposed algorithm and two conventional algorithms were used to cluster three categorical data sets. The proposed method was found to give markedly better clustering results.