• Korean Journal of Cognitive Science
• /
• v.23 no.3
• /
• pp.295-322
• /
• 2012

In this paper, we cluster similar word senses applying vector space model and HAL (Hyperspace Analog to Language). HAL measures corelation among words through a certain size of context (Lund and Burgess 1996). The similarity measurement between a word pair is cosine similarity based on the vector space model, which reduces distortion of space between high frequency words and low frequency words (Salton et al. 1975, Widdows 2004). We use PCA (Principal Component Analysis) and SVD (Singular Value Decomposition) to reduce a large amount of dimensions caused by similarity matrix. For sense similarity clustering, we adopt supervised and non-supervised learning methods. For non-supervised method, we use clustering. For supervised method, we use SVM (Support Vector Machine), Naive Bayes Classifier, and Maximum Entropy Method.

• Journal of Environmental Science International
• /
• v.26 no.9
• /
• pp.999-1011
• /
• 2017

In this study, the Chiu-2D velocity-flow rate distribution based on theoretical background of the entropy probability method was applied to actual ADCP measurement data of Gangjung Stream in Jeju from July 2011 to June 2015 to predict the parameter that take part in velocity distribution of the stream. In addition, surface velocity measured by SIV (Surface Image Velocimeter) was applied to the predicted parameter to calculate discharge. Calculated discharge was compared with observed discharge of ADCP observed during the same time to analyze propriety and applicability of depth of water velocity average conversion factor. To check applicability of the predicted stream parameter, surface velocity and discharge were calculated using SIV and compared with velocity and flow based on ADCP. Discharge calculated by applying velocity factor of SIV to the Chiu-2D velocity-flow rate distribution and discharge based on depth of water velocity average conversion factor of 0.85 were $0.7171m^3/sec$ and $0.5758m^3/sec$, respectively. Their error rates compared to average ADCP discharge of $0.6664m^3/sec$ were respectively 7.63% and 13.64%. Discharge based on the Chiu-2D velocity-flow distribution showed lower error rate compared to discharge based on depth of water velocity average conversion factor of 0.85.

• Journal of Technology Innovation
• /
• v.5 no.1
• /
• pp.15-43
• /
• 1997

This paper investigates empirically the relationship between various business portfolio properties (particularly technological properties) and chaebol's performance using data on the 50largest chaebols in Korea. In addition to the traditional indexes to measure diversification such as entropy index, we calculated inter-industry technological similarity using R'||'&'||'D expenditure data by industry and 1990 Input-output Table in korea, and obtained chaebol-level technological relatedness and internal transaction proportion from chaebols' business profile, inter-inustry technological similarity and 1990 input-output table. We applied factor analysis on 13 business portfolio property indexes and showed that they could be grouped into 3 dimensions. diversification scope, inter-business relatedness and degree of vertical integration. In this paper, using 50 largest chaebols' financial data (1989-1994), we analyzed empirically the effect of business portfolio properties on ROS(Return On Sales) which is conventional index for firm performance and on TFP(Total Factor Productivity) growth which is a pure measure of firm performance. To utilize the advantage of panel data, FEM(Fixed Effect Model) and REM(Random Effect Model) were used. The empirical result shows that the entropy index as a measurement of inter-business relatedness in not significant but technological relatedness index is significant. OLS estimates on pooled data were considerably different from FEM or REM estimates on panel data. By introducing interaction effect among the three variables for business portfolio properties, we obtained three findings. First, only VI(Vertical integration) has a significant positive correlation with ROS. Second, when using TFP growth as an dependent variable, both TR(Technological Relatedness) and VI are significant and positively related to the dependent variable. Third, the interaction term between TR and VI is significant and negatively affects TFP growth, meaning that TR and VI are substitutes. These results suggest strategic directions on restructuring business portfolio. As VI is increased, chaebols will get more profit. A higher level of either TR or VI will increase TFP growth rate, but increase in both TR and VI will have a negative effect on TFP growth. To summarize, certain business portfolio properties such as VI and TR can be considered "resources" themselves since they can affect profit rate and productivity growth. VI and TR have a synergy effect of change in profit rate and productivity growth. VI increases ROS and productivity growth, while TR increases productivity growth representing a technological synergy effect.t.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.25 no.8
• /
• pp.1046-1052
• /
• 2021

Artificial intelligence is being applied in various industrial fields to the development of the fourth industry and the construction of high-performance computing environments. In the medical field, deep learning learning such as cancer, COVID-19, and bone age measurement was performed using medical images such as X-Ray, MRI, and PET and clinical data. In addition, ICT medical fusion technology is being researched by applying smart medical devices, IoT devices and deep learning algorithms. Among these techniques, medical image-based deep learning learning requires accurate finding of medical image biomarkers, minimal loss rate and high accuracy. Therefore, in this paper, we would like to compare and analyze the performance of the Cross-Entropy function used in the image classification algorithm of the loss function that derives the loss rate in the chest X-Ray image-based deep learning learning process.

• Archives of Pharmacal Research
• /
• v.7 no.1
• /
• pp.47-52
• /
• 1984

Four polymorphic forms (I, II, III and IV) of hydrochlorothiazide have been characterized on the basis of x-ray diffractometry and differential thermal analysis. Form I was obtained by crystallization from N, N-dimethylformamide and Form II was crystallized from hot methanol. Form III was precipitated from sodium hydroxide aqueous solution by treatment with hydrochloric acid and Form IV was crystallized from 50% methanol. The metastable form I was a most stable form among four polymorphs, which was stable more than ten months at room temperature. The thermodynamic parameters such as heat of solution, enthalpy, entropy, free energy difference and transition temperature were determined by the measurement of intrinsic dissolution rate. The transition temperature and the heat of transition between the metastable Form I an Form II were determined to be $299.15^{\circ}$K and 5.03 Kcal/mole, respectively and free energy difference ($\delta$ F) was 302. 13 cal/mole. Diuretic action of these four polymorphic forms was also evaluated by monitoring the difference in urinary excretion of sodium, potassium and magnesium in rats.

• Journal of The Korean Astronomical Society
• /
• v.35 no.2
• /
• pp.67-73
• /
• 2002

We have made a topological study of cosmic microwave background (CMB) polarization maps by simulating the AMiBA experiment results. A ACDM CMB sky is adopted to make mock interferometric observations designed for the AMiBA experiment. CMB polarization fields are reconstructed from the AMiBA mock visibility data using the maximum entropy method. We have also considered effects of Galactic foregrounds on the CMB polarization fields. The genus statistic is calculated from the simulated Q and U polarization maps, where Q and U are Stokes parameters. Our study shows that the Galactic foreground emission, even at low Galactic latitude, is expected to have small effects on the CMB polarization field. Increasing survey area and integration time is essential to detect non-Gaussian signals of cosmological origin through genus measurement.

• Proceedings of the Korean Information Science Society Conference
• /
• 2001.10b
• /
• pp.115-117
• /
• 2001

기존의 인터넷 웹사이트에서는 사용자의 만족을 극대화시키기 위하여 사용자별로 개인화 된 서비스를 제공하는 협력적 필터링 방식을 적용하고 있다. 협력적 필터링 기술은 사용자의 취향에 맞는 아이템을 예측하여 추천하며, 비슷한 선호도를 가진 다른 사용자들과의 상관관계를 구하기 위하여 일반적으로 피어슨 상관계수를 많이 이용한다. 그러나, 피어슨 상관계수를 이용한 방법은 사용자가 평가를 한 아이템이 있을 때에만 상관관계를 구할 수 있다는 단점과 예측의 정확성이 떨어진다는 단점을 가지고 있다. 따라서, 본 논문에서는 피어슨 상관관계 기반 예측 기법을 보완하여 보다 정확한 사용자 유사도를 구하는 방법을 제안한다. 제안된 방법에서는 사용자들을 대상으로 사용자가 평가를 한 아이템의 선호도를 사용해서 엔트로피를 적용하였고, 사용자가 선호도를 표시하지 않은 상품에 대해서는 Default Voting 방법을 이용하여 보다 정확한 헙력적 필터링 방식을 구현하였다.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 2001.04a
• /
• pp.37-40
• /
• 2001

This paper describes the selection and evaluation of RP processes. Major rapid prototyping processes such as SLS, SLA, FDM and LOM, which are wide spread in use are selected. A test part, which includes various primitives, is designed in order to evaluate these RP processes. Measurement of the test part is automated by using a CMN program. To visualize and analyze measured data, Microsoft Access and Visual C++ are used. Also, from measured data obtained, TOPSIS, one of the decision making methods, and Shannon Entropy is used to select an appropriate RP process for specific application.

• Korean Journal of Air-Conditioning and Refrigeration Engineering
• /
• v.16 no.4
• /
• pp.383-389
• /
• 2004

Magnetic refrigeration is based on the magnetocaloric effect (MCE) - the ability of some materials to heat up when magnetized and cool down when removed from the magnetic field. The available techniques for studying the MCE we: (1) direct measurements by monitoring the change in the material's temperature during the application or removal of the magnetic field; and (2) indirect calculations from the experimental data of magnetization and/or specific heat as a function of the temperature and magnetic field. The MCE of gadolinium (Gd) has been demonstrated by direct measurements of temperature change, and isothermal magnetic entropy changes and adiabatic temperature changes have been calculated.

• Journal of Biomedical Engineering Research
• /
• v.11 no.2
• /
• pp.243-248
• /
• 1990

EMG (eleltromyographic) signals are generated by contracting muscle and detected in and out side of muscle in the form of random signals. In the measurement of muscle fatigue, the mean frequency of EMG signals using spectrum analysis is an important parameter in diagonosis of muscle disease and in sports medicine fields. In this study, the degree of spectral transfer to lower frequency caused by accumulation of Latic acid inside the muscle is estimated. The new spectral analysis method using 2"d order hAaximum Entropy Method was applied to estimate the mean frequency and we confirmed that this new method yields fast and reliable estimation.tion.