• Proceedings of the Korean Society for Bioinformatics Conference
• /
• 2005.09a
• /
• pp.124-129
• /
• 2005

The Growing Self-Organizing Map (GSOM), an extended type of the Self-Organizing Map, is a widely accepted tool for clustering high dimensional data. It is also suitable for the clustering of short DNA sequences of phylogenetic genomes by their oligonucleotide frequency. The GSOM presents the result of the clustering process visually on a coloured map, where the clusters can be identified by the user. This paper describes a proposal for automatic cluster detection on this map without any participation by the user. It has been applied with good success on 20 different data sets for the purpose of species separation.

• Communications for Statistical Applications and Methods
• /
• v.8 no.3
• /
• pp.643-650
• /
• 2001

Some classical plots like scatterplot matrices and parallel coordinates are valuable tools for data visualization. These tools are extensively used in the modern data mining softwares to explore the inherent data structure, and hence to visually classify or cluster the database into appropriate groups. However, the interpretation of these plots are very sensitive to the arrangement of variables. In this work, we introduce two methods to arrange the variables for data visualization. First method is based on the work of Wegman (1999), and this is to arrange the variables using minimum distance among all the pairwise permutation of the variables. Second method is using the idea of principal components. We Investigate the effectiveness of these methods with parallel coordinates using real data sets, and show that each of the two proposed methods has its own strength from different aspects respectively.

• 제어로봇시스템학회:학술대회논문집
• /
• 1993.10a
• /
• pp.660-665
• /
• 1993

Neural networks are used in the framework of sensorbased tracking control of robot manipulators. They learn by practice movements the relationship between PSD ( an analog Position Sensitive Detector) sensor readings for target positions and the joint commands to reach them. Using this configuration, the system can track or follow a moving or stationary object in real time. Furthermore, an efficient neural network architecture has been developed for real time learning. This network uses multiple sets of simple backpropagation networks one of which is selected according to which division (corresponding to a cluster of the self-organizing feature map) in data space the current input data belongs to. This lends itself to a very fast training and processing implementation required for real time control.

• International Journal of Human Ecology
• /
• v.7 no.1
• /
• pp.77-83
• /
• 2006

The ability to customize garments for fit in the apparel industry is directly tied to the availability of comprehensive and accurate sets of anthropometrical data for each consumer. The data for apparel sizing systems is available from three major standard/ research organizations: ASTM (American Society for Testing and Materials), ISO (International Standard Organization), and NCHS (National Center for Health Statistics). However, these standards ignore various head shapes and are outdated for the development future head products. This creates a data gap an ever changing multi-cultural society such as the United Sates. Although major government and industry safety organizations recognize the importance of safety for head products, few studies were found to support their reasoning. The purpose of this study is to provide accurate head dimension data for developing safety head products by analyzing various head shapes in the United Sates which includes various ethnic backgrounds. This study was carried out on 105 males in the United States. Factor analysis, cluster analysis, Moreover, Duncan analysis were all used for analyzing various head shapes.

• Proceedings of the Korean Society of Machine Tool Engineers Conference
• /
• 2001.04a
• /
• pp.239-243
• /
• 2001

Neural networks are used in the framework of sensorbased tracking control of robot manipulators. They learn by practice movements the relationship between PSD (an analog Position Sensitive Detector) sensor readings for target positions and the joint commands to reach them. Using this configuration, the system can track or follow a moving or stationary object in real time. Furthermore, an efficient neural network architecture has been developed for real time learning. This network uses multiple sets of simple backpropagation networks one of which is selected according to which division (corresponding to a cluster of the self-organizing feature map) in data space the current input data belongs to. This lends itself to a very training and processing implementation required for real time control.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 2003.05a
• /
• pp.729-733
• /
• 2003

The presence of local features within clusters incurred by multi-modal nature of data prohibits many conventional clustering techniques from working properly. Especially, the clustering of datasets with non-Gaussian distributions within a cluster can be problematic when the technique with implicit assumption of Gaussian distribution is used. Current study proposes a simple tandem clustering method composed of k-means type algorithm and hierarchical method to solve such problems. The multi-modal dataset is first divided into many small pre-clusters by k-means or fuzzy k-means algorithm. The pre-clusters found from the first step are to be clustered again using agglomerative hierarchical clustering method with Kullback- Leibler divergence as the measure of dissimilarity. This method is not only effective at extracting the multi-modal clusters but also fast and easy in terms of computation complexity and relatively robust at the presence of outliers. The performance of the proposed method was evaluated on three generated datasets and six sets of publicly known real world data.

• Asia pacific journal of information systems
• /
• v.15 no.1
• /
• pp.25-44
• /
• 2005

This study examines control types of IS outsourcing project management focusing on two sets of questions: What control types are implemented during IS outsourcing projects? How various control types affect the performance of IS outsourcing projects. In order to meet these objectives, four control types are developed using typology-approach: dynamic, contract-oriented, partnership-oriented, and passive. And then four control types are identified based on taxonomy-approach using cluster analysis from 66 projects: dynamic, contract-oriented, passive, and middle. The result shows that the derived taxonomy-types are similar to the typology-types except partnership-oriented type. The result also indicates that dynamic and contract-oriented types are in the highest performance, while passive type is in the least performance.

• Proceedings of the Korean Society of Machine Tool Engineers Conference
• /
• 2000.04a
• /
• pp.171-175
• /
• 2000

Neural networks are in the framework of sensorbased tracking control of robot manipulators. They learn by practice movements the relationship between PSD (an analog Position Sensitive Detector) sensor readings for target positions and the joint commands to reach them. Using this configuration, the system can track or follow a moving or stationary object in real time. Furthermore, an efficient neural network architecture has been developed for real time learning. This network uses multiple sets of simple backpropagation networks one of which is selected according to which division (corresponding to a cluster of the self-organizing feature map) in data space the current input data belongs to. This lends itself to a very training and processing implementation required for real time control.

• Journal of the Korean Mathematical Society
• /
• v.49 no.1
• /
• pp.17-30
• /
• 2012

Let D be an arbitrary domain in $\mathbb{C}^n$, n > 1, and $M{\subset}{\partial}D$ be an open piece of the boundary. Suppose that M is connected and ${\partial}D$ is smooth real-analytic of finite type (in the sense of D'Angelo) in a neighborhood of $\bar{M}$. Let f : $D{\rightarrow}\mathbb{C}^n$ be a holomorphic correspondence such that the cluster set $cl_f$(M) is contained in a smooth closed real-algebraic hypersurface M' in $\mathbb{C}^n$ of finite type. It is shown that if f extends continuously to some open peace of M, then f extends as a holomorphic correspondence across M. As an application, we prove that any proper holomorphic correspondence from a bounded domain D in $\mathbb{C}^n$ with smooth real-analytic boundary onto a bounded domain D' in $\mathbb{C}^n$ with smooth real-algebraic boundary extends as a holomorphic correspondence to a neighborhood of $\bar{D}$.

• Korean Journal of Organic Agriculture
• /
• v.28 no.2
• /
• pp.189-208
• /
• 2020

The purpose of this study aims to estimate consumption selection attribute, part-worth of organic instant rice through the use of conjoint analysis method. The conjoint analysis is to trace the development of consumer preference among multi-attribute alternatives. The selection attribute was including 4 factors preferred Type of rice, Capacity, Brand and payment price. For this research, a total of 192 questionnaires was collected of which 200 were completed. The research design was a full profile method by orthogonal design then 9 main profiles, 3 holdout sets were created. The results of this research were as follows. Consumers of organic instant rice are consider their importance of selection attributes was in order to price (25.87%), Type of rice (27.231%), Brand/Purchase channel (24.013%) and Capacity (18.494%). The findings of this study have identified 3 clusters for each experience visitors. Each cluster has a different and showed the relative importance or preference values for each accessible attribute of the segmentation.