• Communications for Statistical Applications and Methods
• /
• v.29 no.2
• /
• pp.193-202
• /
• 2022

Semi-supervised learning has gained significant attention in recent applications. In this article, we provide a selective overview of popular semi-supervised methods and then propose a simple but effective algorithm for semi-supervised classification using support vector machines (SVM), one of the most popular binary classifiers in a machine learning community. The idea is simple as follows. First, we apply the dimension reduction to the unlabeled observations and cluster them to assign labels on the reduced space. SVM is then employed to the combined set of labeled and unlabeled observations to construct a classification rule. The use of SVM enables us to extend it to the nonlinear counterpart via kernel trick. Our numerical experiments under various scenarios demonstrate that the proposed method is promising in semi-supervised classification.

• The KIPS Transactions:PartB
• /
• v.15B no.4
• /
• pp.285-292
• /
• 2008

Color histograms have been used as feature vectors to characterize the color features of given images, but they have a limitation in efficiency by generating high-dimensional feature vectors. In this paper, we present a method to reduce the dimension of the feature vectors by applying PCA (principal components analysis) to the color histogram of a given vehicle image. With SVM (support vector machine) method, the dimension-reduced feature vectors are used to recognize the colors of vehicles. After reducing the dimension of the feature vector by a factor of 32, the successful recognition rate is reduced only 1.42% compared to the case when we use original feature vectors. Moreover, the computation time for the color recognition is reduced by a factor of 31, so we could recognize the colors efficiently.

• Journal of Korea Society of Digital Industry and Information Management
• /
• v.15 no.4
• /
• pp.41-49
• /
• 2019

Patents are system to promote the development of industry by disclosing technology. The importance of recent patent is being emphasized. For this reason, companies apply for many patents. And they analyze the patent. Patent analysis helps to protect and foster their technology. Previously this method has been carried out by experts. Expert-based patent analysis, however, has the disadvantage of being time-consuming and expensive. Consequently, we try to solve this problems by developing prediction model. Therefore, this paper proposes a data-based patent analysis method using quantitative indicator and textual information. We confirmed the practical applicability of the proposed method through 1,831 autonomous vehicle patents. As a result, it was possible to confirmed that safety and lane detection related technologies are important.

• Asian Journal of Atmospheric Environment
• /
• v.11 no.4
• /
• pp.283-299
• /
• 2017

An analytical method using high-performance liquid chromatography (HPLC) with fluorescence (FL) detection was developed for simultaneously analyzing 10 polycyclic aromatic hydrocarbons (PAHs) and 18 nitro-derivatives of PAHs (NPAHs). The two-dimensional HPLC system consists of an on-line clean-up and reduction for NPAHs in the 1st dimension, and separation of the PAHs and the reduced NPAHs and their FL detection in the 2nd dimension after column-switching. To identify an ideal clean-up column for removing sample matrix that may interfere with detection of the analytes, the characteristics of 8 reversed-phase columns were evaluated. The nitrophenylethyl (NPE)-bonded silica column was selected because of its shorter elution band and larger retention factors of the analytes due to strong dipole-dipole interactions. The amino-substituted PAHs (reduced NPAHs), PAHs and deuterated internal standards were separated on polymeric octadecyl-bonded silica (ODS) columns and by dual-channel detection within 120 min including clean-up and reduction steps. The limits of detection were 0.1-9.2 pg per injection for PAHs and 0.1-140 pg per injection for NPAHs. For validation, the method was applied to analyze crude extracts of fine particulate matter ($PM_{2.5}$) samples and achieved good analytical precision and accuracy. Moreover, the standard reference material (SRM1649b, urban dust) was analyzed by this method and the observed concentrations of PAHs and NPAHs were similar to those in previous reports. Thus, the method developed here-in has the potential to become a standard HPLC-based method, especially for NPAHs.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2001.11a
• /
• pp.127-132
• /
• 2001

A successful design approach for noise barriers should be multidisciplinary because noise reduction goals influence both acoustical and non-acoustical considerations, such as maintenance, safety, physical construction, cost, and visual impact. These various barrier design options are closely related with barrier dimensions. In this study, we have proposed an optimal design method of straight noise barriers using genetic algorithm, providing a barrier having the smallest dimension and achieving the specified noise reduction at a receiver region exposed to the industry and traffic noise, to help a successful barrier design.

• Communications of the Korean Mathematical Society
• /
• v.15 no.1
• /
• pp.133-142
• /
• 2000

We consider some numerical solution methods for equilibrium equations Af + E$^{T}$ λ = r, Ef = s. Algebraic problems of this form evolve from many applications such as structural optimization, fluid flow, and circuits. An important approach, called the force method, to the solution to such problems involves dimension reduction nullspace computation for E. The purpose of this paper is to investigate the substructuring method for the solution step of the force method in the context of the incompressible fluid flow. We also suggests some iterative methods based upon substructuring scheme..

• Proceedings of the Korean Statistical Society Conference
• /
• 2003.10a
• /
• pp.9-14
• /
• 2003

A Bayesian model-based clustering method is proposed for clustering objects on the basis of dissimilarites. This combines two basic ideas. The first is that tile objects have latent positions in a Euclidean space, and that the observed dissimilarities are measurements of the Euclidean distances with error. The second idea is that the latent positions are generated from a mixture of multivariate normal distributions, each one corresponding to a cluster. We estimate the resulting model in a Bayesian way using Markov chain Monte Carlo. The method carries out multidimensional scaling and model-based clustering simultaneously, and yields good object configurations and good clustering results with reasonable measures of clustering uncertainties. In the examples we studied, the clustering results based on low-dimensional configurations were almost as good as those based on high-dimensional ones. Thus tile method can be used as a tool for dimension reduction when clustering high-dimensional objects, which may be useful especially for visual inspection of clusters. We also propose a Bayesian criterion for choosing the dimension of the object configuration and the number of clusters simultaneously. This is easy to compute and works reasonably well in simulations and real examples.

• The Journal of the Acoustical Society of Korea
• /
• v.37 no.6
• /
• pp.518-524
• /
• 2018

The cover song refers to live recordings or reproduced albums. This paper studies two-dimensional Fourier transform as a feature-dimension reduction method to search cover song fast. The two-dimensional Fourier transform is conducive in feature-dimension reduction for cover song search due to musical-key invariance. This paper extends the previous work, which only utilize the magnitude of the Fourier transform, by introducing an invariant from phase based on the assumption that adjacent frames have the same musical-key change. We compare the cover song retrieval accuracy of the Fourier-transform based methods over two datasets. The experimental results show that the addition of the invariant from phase improves the cover song retrieval accuracy over the previous magnitude-only method.

• Tunnel and Underground Space
• /
• v.22 no.2
• /
• pp.120-129
• /
• 2012

Recently, mining facilities have being installed in an underground space according to a social demand for environment-friendly mine development. The underground structures for mining facilities usually requires a large volume of space with width greater than height, and thus the stability assessment of the large-scale underground mine structure is an important issue. In this study, we analysed a factor of safety based on strength reduction method, and proposed a numerical design approach to determining the dimension of underground mine structures in combination with a strength reduction method and a multivariate regression analysis. Input design parameters considered in the present study were the stress ratio and shear strength of rock mass, and the width and cover depth of underground mine structures. The stabilities of underground mine structures were assessed in terms of factor of safety under different conditions of the above input parameters. It was calculated by the strength reduction method, and several kinds of fit functions were obtained through various multivariate regression analyses. Using a best-fit regression model, we proposed the charts which provide preliminary design information on the dimension of underground mine structures.

• ETRI Journal
• /
• v.33 no.4
• /
• pp.652-655
• /
• 2011

In this letter, we propose a new compression method for a high dimensional support vector machine (SVM). We used singular value decomposition (SVD) to compress the norm part of a radial basis function SVM. By deleting the least significant vectors that are extracted from the decomposition, we can compress each vector with minimized energy loss. We select the compressed vector dimension according to the predefined threshold which can limit the energy loss to design criteria. We verified the proposed vector compressed SVM (VCSVM) for conventional datasets. Experimental results show that VCSVM can reduce computational complexity and memory by more than 40% without reduction in accuracy when classifying a 20,958 dimension dataset.