• The Research Journal of the Costume Culture
• /
• v.15 no.3 s.68
• /
• pp.525-540
• /
• 2007

This study intended to construct the brand switching matrix in the Korean casual wear market and to analyze it in various aspects. 1,014 sample data were collected in Seoul area, a center of fashion retailing. Since the respondents cited over 200 brand names as their last 2 purchased casual wear brands, 15 most frequently-purchased brands were selected for constructing the brand switching matrix. As a result of the examination, it was founded that the brand loyalty was dominant rather than brand switching in the casual wear market. Polo was identified as the leading brand in the market. Its brand equity, which was comprised of brand recognition, brand preference (loyalty), perceived quality, and brand association, was evaluated very high. Especially, the strength of Polo was the consumer's strong preference and the brand image of simplicity, naturalness, and neatness. After combining 15 brands into 6 groups based on the style and price, additional interpretation was performed on this 'trend switching matrix.' A transition of fashion trend in casual wear was observed. Applying the brand switching matrix on fashion products gave us much insight to the market.

• Journal of the Korean Data and Information Science Society
• /
• v.21 no.3
• /
• pp.547-554
• /
• 2010

In this paper, we extend the work by Lee et al. (2010) to multidimensional di usion processes. A test statistic analogous to the one-dimensional case is proposed to inves-tigate the joint stability of covariance matrix parameters and, under certain regularity conditions, is shown to have a limiting distribution of the sup of a multidimensional Brownian bridge. A simulation result is provided for illustration.

• Proceedings of the Computational Structural Engineering Institute Conference
• /
• 1990.04a
• /
• pp.24-30
• /
• 1990

An improved algorithm for simulation of multi-variate random processes has been presented. It is based on the spectral representation method. The conventional methods give sample time histories which satisfy the target spectral density matrix only in the sense of ensemble average. However, the present method can generate sample functions which satisfy the target spectra in the ergodic sense. Example analysis is given for the simulation of earthquake accelerations with three components.

• KIEE International Transactions on Electrophysics and Applications
• /
• v.5C no.6
• /
• pp.251-256
• /
• 2005

Directionally melt-textured high $T_c\;(Y_{0.5}Nd_{0.25}Sm_{0.25})Ba_{2}Cu_{3}O_y$ [(YNS)-123] superconductor was systematically investigated by the zone melt growth process in air. A sample prepared by this method showed well-textured microstructure, and $(Y_{0.5}Nd_{0.25}Sm_{0.25})_{2}BaCuO_5$[(YNS)211] inclusions were uniformly dispersed in large $(Y_{0.5}Nd_{0.25}Sm_{0.25})Ba_{2}Cu_{3}O_y$ [(YNS)123] matrix. High irreversibility field and magnetization hysteresis loop of the zone melt-textured (YNS)-123 sample exhibited the enhanced flux pinning, compared with $YBa_{2}Cu_{3}O_y$ (Y-123) sample without RE(rare earth). Critical current density of (YNS)-123 sample was $2.5{\times}10^4\;A/cm^2$ at 2 T and 77 K.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.7 no.11
• /
• pp.2657-2675
• /
• 2013

A novel neighbor selection-based fingerprinting algorithm using matrix correlation (MC) for Wi-Fi localization is presented in this paper. Compared with classic fingerprinting algorithms that usually employ a single received signal strength (RSS) sample, the presented algorithm uses multiple on-line RSS samples in the form of a matrix and measures correlations between the on-line RSS matrix and RSS matrices in the radio-map. The algorithm makes efficient use of on-line RSS information and considers RSS variations of reference points (RPs) for localization, so it offers more accurate localization results than classic neighbor selection-based algorithms. Based on the MC algorithm, an error estimation method using artificial neural network is also presented to fuse available information that includes RSS samples and localization results computed by the MC algorithm and model the nonlinear relationship between the available information and localization errors. In the on-line phase, localization errors are estimated and then used to correct the localization results to reduce negative influences caused by a static radio-map and RP distribution. Experimental results demonstrate that the MC algorithm outperforms the other neighbor selection-based algorithms and the error estimation method can reduce the mean of localization errors by nearly half.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.12 no.1
• /
• pp.368-391
• /
• 2018

Face recognition (FR) with a single sample per person (SSPP) is common in real-world face recognition applications. In this scenario, it is hard to predict intra-class variations of query samples by gallery samples due to the lack of sufficient training samples. Inspired by the fact that similar faces have similar intra-class variations, we propose a virtual sample generating algorithm called k nearest neighbors based virtual sample generating (kNNVSG) to enrich intra-class variation information for training samples. Furthermore, in order to use the intra-class variation information of the virtual samples generated by kNNVSG algorithm, we propose image set based multimanifold discriminant learning (ISMMDL) algorithm. For ISMMDL algorithm, it learns a projection matrix for each manifold modeled by the local patches of the images of each class, which aims to minimize the margins of intra-manifold and maximize the margins of inter-manifold simultaneously in low-dimensional feature space. Finally, by comprehensively using kNNVSG and ISMMDL algorithms, we propose k nearest neighbor virtual image set based multimanifold discriminant learning (kNNMMDL) approach for single sample face recognition (SSFR) tasks. Experimental results on AR, Multi-PIE and LFW face datasets demonstrate that our approach has promising abilities for SSFR with expression, illumination and disguise variations.

• Analytical Science and Technology
• /
• v.11 no.2
• /
• pp.112-119
• /
• 1998

A Siemens SemiQuant (SSQ) 3000 program, a precalibrated 'standardless' analytical program handling up to 90 elements, was evaluated for the fast analysis of various types of reference materials using a wavelength dispersive X-ray spectrometer. Various types of standard reference materials such as metal discs, metal chips, and geological materials in powder form were analysed and it took 23 minutes of measuring time for 75 elements. Measurements of geological reference materials using different sampling methods were carried out and their data were interactively evaluated. The analysis of materials of a known matrix concentration such as stainless steels provided higher precision value compared to totally unknown samples. The analyses of materials prepared as pressed pellets or fused glass beads provided higher precision values compared to the measurement of loose powders with a foil on the sample surface and helium operation, though their sampling procedures were more complicate and took more time. Since very light elements such as boron, carbon, and oxygen have a strong influence on the matrix effects and also on the calculation of effective matrix corrections, the rhodium Compton check was applied to verify the reliability of the defined light element concentrations of light matrix materials and the defined major sample compounds. Failure of defining correct matrix resulted in an unoptimized matrix correction and therefore in the wrong calculation of the element concentration.

• Bulletin of the Korean Chemical Society
• /
• v.33 no.3
• /
• pp.987-992
• /
• 2012

Magnetic nanoparticles (MNPs) treated with phosphoric acid were used to improve sequence coverage in protein identification by matrix-assisted laser desorption/ionization mass spectrometry (MALDI MS). Sample solution of tryptic peptides from proteins was mixed with the MNPs, and the MNPs were separated from the supernatant using a magnet. MALDI mass spectra obtained separately from the supernatant and the MNPs were distinctly different and complementary to each other. Combination of the two spectra led to a significantly increased sequence coverage.

• MALSORI
• /
• v.55
• /
• pp.93-102
• /
• 2005

In this paper, sub-stream based eigenvoice method is proposed to overcome the weak points of conventional eigenvoice and dimensional eigenvoice. In the proposed method, sub-streams are automatically constructed by the statistical clustering analysis that uses the correlation information between dimensions. To obtain the reliable distance matrix from covariance matrix for dividing into optimal sub-streams, MAP adaptation technique is employed to the covariance matrix of training data and the sample covariance of adaptation data. According to our experiments, the proposed method shows $41\%$ error rate reduction when the number of adaptation data is 50.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2014.06a
• /
• pp.230-231
• /
• 2014

Compressive sensing is a new sampling technique, which allows to sample a signal under the Nyquist-Shannon sampling rate. For block-based compressive sensing, a hybrid sensing matrix which contains low-frequency patterns in addition to the random Gaussian numbers is good for exploiting typical property of natural images. By noting that MH-BCS-SPL is well known for its good recovery performance, this paper investigates effect of the hybrid sensing matrix on MH-BCS-SPL in the sense of how large portion of low-frequency patterns can provide performance improvement.