• The Korean Journal of Applied Statistics
• /
• v.21 no.1
• /
• pp.141-149
• /
• 2008

Recently we frequently analyze multivariate data with quite large number of variables. In such data sets, virtually duplicated variables may exist simultaneously even though they are conceptually distinguishable. Duplicate variables may cause problems such as the distortion of principal axes in principal component analysis and factor analysis and the distortion of the distances between observations, i.e. the input for cluster analysis. Also in supervised learning or regression analysis, duplicated explanatory variables often cause the instability of fitted models. Since real data analyses are aimed often at multiple purposes, it is necessary to reduce the number of variables to a parsimonious level. The aim of this paper is to propose a practical algorithm for selection of a subset of variables from a given set of p input variables, by the criterion of minimum trace of partial variances of unselected variables unexplained by selected variables. The usefulness of proposed method is demonstrated in visualizing the relationship between selected and unselected variables, in building a predictive model with very large number of independent variables, and in reducing the number of variables and purging/merging categories in categorical data.

• Korean Journal of Cognitive Science
• /
• v.19 no.1
• /
• pp.75-88
• /
• 2008

Visual scene contains lots of redundant information. To process this redundant information without increasing brain's volume, human visual system may summarize incoming information. If similar but different information are given to visual system, visual system extracts statistical properties of the information. One example of the statistical representation is representation of mean size. The mean representation is accurate and durable. The process of mean representation is suggested to be parallel. However, previous studies on the mean representation mostly used behavioral methods. The purpose of this study was to investigate which neural regions extracted the mean size of a set of circles using fMRI method. According to previous studies, BOLD signal of certain areas that were in charge of cousin stimuli decreased when the same stimuli presented repetitively. We used this paradigm and found that BOLD signal of right occipital area was decreased when same mean site was presented repeatedly. This results suggest that right occipital area is the locus of mean representation of visual stimuli.

• Journal of the Korea Society of Computer and Information
• /
• v.20 no.4
• /
• pp.95-102
• /
• 2015

This paper proposes a simple algorithm for circuit minimization. There are currently two effective heuristics for circuit minimization, namely manual Karnaugh maps and computable Quine-McCluskey algorithm. The latter, however, has a major defect: the runtime and memory required grow $3^n/n$ times for every increase in the number of variables n. The proposed algorithm, however, extracts the prime implicants (PI) that cover minterms of a given Boolean function by deriving an implicants table based on frequency. From a set of the extracted prime implicants, the algorithm then eliminates redundant PIs again based on frequency. The proposed algorithm is therefore capable of minimizing circuits polynomial time when faced with an increase in n. When applied to various 3-variable and 4-variable cases, it has proved to swiftly and accurately obtain the optimal solutions.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.20 no.9
• /
• pp.1723-1729
• /
• 2016

Digital watermarking is tried to get an optimum tradeoffs between its performance characteristics, robustness, transparency and capacity. This paper is, therefore, suggesting a watermarking technique that builds multiple barcodes in various frequency bands to implement embedded watermarks that is imperceptible and robust against various attacks. Even though a watermark technique with duplicated barcode watermarks embedded in various frequency bands can satisfy robustness as there is high possibility that watermarks embedded in an image remains after various attacks, the duplicated barcode data can weaken imperceptibility. Thus, to satisfy the conflicting characteristic requirements of watermarks, robustness and imperceptibility, different barcode data is embedded in each frequency band. The test shows that ownership authentication with the technique suggested in this thesis does not require specialized hardware, and extracted watermarks can be easily identified through a mobile barcode scanner app, which allows low complexity, low cost and swift identification.

• Journal of Information Management
• /
• v.41 no.3
• /
• pp.155-173
• /
• 2010

With the advanced learning and the diversity of topics, researchers on each area keenly feel the need of precise and a quick discovery of required information at any time. This study presents a way of constructing the automatic mapping system that can compare and analyze duplicated data and that describes the result by building an effective reference extraction method and another way of correcting the wrong form of used Chinese characters with Traditional Korean Medicine dictionary. With this innovation, data duplication on references and Chinese characters errors can be fixed. Under the situation that a number of references of newly published papers that can continuously be extracted.

• Journal of Digital Policy
• /
• v.2 no.3
• /
• pp.1-7
• /
• 2023

In this paper, I proposes a method for performing single image super resolution by separating texture-spatial domains and then classifying features based on detailed information. In CNN (Convolutional Neural Network) based super resolution, the complex procedures and generation of redundant feature information in feature estimation process for enhancing details can lead to quality degradation in super resolution. The proposed method reduced procedural complexity and minimizes generation of redundant feature information by splitting input image into two channels: texture and spatial. In texture channel, a feature refinement process with step-wise skip connections is applied for detail restoration, while in spatial channel, a method is introduced to preserve the structural features of the image. Experimental results using proposed method demonstrate improved performance in terms of PSNR and SSIM evaluations compared to existing super resolution methods, confirmed the enhancement in quality.

• Proceedings of the KAIS Fall Conference
• /
• 2010.05b
• /
• pp.1197-1200
• /
• 2010

Post-genomic 시대에 접어들면서 단백질의 기능의 주석이 중요한 문제로 떠오르기 시작하였다. 이런 단백질 기능을 예측하기 위해 단백질 상호작용(Protein-Protein interaction) 데이터를 이용한 방법들이 지난 10여 년간 발표되어왔다. 단백질 상호작용(Protein-Protein interaction) 데이터는 단백질들 간의 서열 등의 특징을 이용해 상호간의 연결 관련성이 있는 단백질끼리의 관계를 네트워크로 나타낸 자료이다. 현재 이러한 단백질 상호작용(Protein-Protein interaction) 데이터들은 MIPS, DIP, BioGrid등 약 5~6군데에서 제공되고 있다. 각각의 데이터는 다른 형식을 가지고 있고, 중복되는 정보도 포함하고 있다. 여러 연구 방법에서 데이터를 사용할 때 한군데에서만 추출하기 보다는 여러 데이터에서 추출하는 경우가 많기 때문에 다른 형식의 데이터를 이용하는데 불필요한 수고가 들어가게 된다. 때문에 여러군데의 데이터를 한 가지 형식으로 맞추어 통합적으로 구축하여 연구 시 데이터 사용에 용이하도록 설계 하였다. 또한 발표된 단백질 기능 예측 방법에 대한 정리를 통해 앞으로의 연구를 하는데 있어서 필요한 자료를 얻고 열람할 수 있도록 설계하였다. 이를 통해 관련 연구를 하거나 관심이 있는 사람들의 데이터를 검색하는데 많은 도움이 될 것이다.

• Journal of the Korean BIBLIA Society for library and Information Science
• /
• v.13 no.1
• /
• pp.157-165
• /
• 2002

The purpose of this study is to achieve two goals. One is to construct knowledge database which could read and analyse electronic documents in place of researchers for the purpose of improvement of research productivity, and the other is to develop automated patent application engine which is connected to the knowledge database. This study discusses the possibilities and appropriateness of two systems mentioned earlier, and provides elements necessary to system developments and technical problems through model development.

• Proceedings of the Korea Institute of Convergence Signal Processing
• /
• 2000.12a
• /
• pp.93-96
• /
• 2000

In this paper, we suggest a feature points selection method using block-based watershed segmentation and polygon approximation for preprocessing of MPEG-4 mesh generation. 2D natural image is segmented by 8$\times$8 or 4$\times$4 block classification method and watershed algorithm. As this result, pixels on the watershed lines represent scene's interior feature and this lines are shapes of closed contour. Continuous pixels on the watershed lines are selected out feature points using Polygon approximation and post processing.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.49 no.2
• /
• pp.113-121
• /
• 2012

In this paper, we proposed fast and efficient algorithm for detecting near-duplication based on content based retrieval in large scale video database. For handling large amounts of video easily, we split the video into small segment using scene change detection. In case of video services and copyright related business models, it is need to technology that detect near-duplicates, that longer matched video than to search video containing short part or a frame of original. To detect near-duplicate video, we proposed motion distribution and frame descriptor in a video segment. The motion distribution descriptor is constructed by obtaining motion vector from macro blocks during the video decoding process. When matching between descriptors, we use the motion distribution descriptor as filtering to improving matching speed. However, motion distribution has low discriminability. To improve discrimination, we decide to identification using frame descriptor extracted from selected representative frames within a scene segmentation. The proposed algorithm shows high success rate and low false alarm rate. In addition, the matching speed of this descriptor is very fast, we confirm this algorithm can be useful to practical application.