• The Journal of the Acoustical Society of Korea
• /
• v.18 no.4
• /
• pp.31-37
• /
• 1999

A bimodal speech recognition using visual and audio information has been proposed and researched to improve the performance of ASR(Automatic Speech Recognition) system in noisy environments. The integration method of two modalities can be usually classified into an early integration and a late integration. The early integration method includes a method using a fixed weight of lip parameters and a method using a variable weight according to speech SNR information. The 4 late integration methods are a method using audio and visual information independently, a method using speech optimal path, a method using lip optimal path and a way using speech SNR information. Among these 6 methods, the method using the fixed weight of lip parameter showed a better recognition rate.

• Proceedings of the Korea Electromagnetic Engineering Society Conference
• /
• 2001.11a
• /
• pp.194-197
• /
• 2001

본 논문에서는 결합창으로 결합을 구현한 유전체 공진기 필터의 결합창 특성을 해석하였다. 결합창의 특성을 결합계수로 표현하였으며, 간단한 표현식으로 유도하였다. 결합창의 크기를 변화시키면서 해석 방법을 적용하여 결합계수를 계산하였고, 측정결과와 비교하여 해석의 유효성을 증명하였다. 본 논문에서 제안한 방법은 적용이 간단할 뿐 아니라, 결합창의 제작 및 측정오차 범위에서 정확한 결과를 보였다. 본 논문에서 제안한 방법은 유전체 공진기 필터의 설계 및 유전체 공진기와 Varactor diode로 이루어지는 동조필터가 동조주파수 대역에서 일정한 통과대역폭을 가지기 위한 최적화된 결합창의 설계적용이 가능하다.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 1994.10a
• /
• pp.238-243
• /
• 1994

본 논문에서는 여러가지 형태의 결합부에 적용할 수 있는 일반적인 모델링 기법에 대하여 기술하였다. 기존의 해석방법은 특정한 결합부에만 적용이 가능하고 유한요소해석과 실험이 상호 보완적인 관계를 가진 형태의 해석방법이므로 실험이 불가능한 경우에 대해서는 적용하기 어려운 경우도 있다. 본 논문에서는 별도의 실험을 수행하지 않고 결합부 영역만을 상세하게 유한요소 모델링을 하여서 선택된 자유도에 대한 유연성 행렬(Flexibility Matrix)을 구하여 결합부의 특성을 구하는 일반적 모델링 기법을 제시하였다. 이 방법은 수치적으로 축약할 수 없는 모델-결합부가 접촉면(Contact Surface)을 가지고 있는 구조물을 효과적으로 축약할 수 있는 장점이 있다. 또한 모델링되는 결합부의 경계조건의 영향을 배제할 수 있으며 결합부에 존재하는 비선형성도 적정범위내에서 선형화할 수 있다. 제시한 일반적 모델링 기법을 나사 결합부, 접착제 결합부(Glued Joints), 볼트 결합부에 적용하여 결합부이 특성을 구하였으며 실험을 통하여 제시한 해석방법의 타당성을 검증하였다.

• Journal of the Mineralogical Society of Korea
• /
• v.16 no.1
• /
• pp.65-73
• /
• 2003

In this study we have derived the two correction methods of hydrogen bonding distance. In case of the intermediate or long hydrogen bond(＞2.5 $\AA$), hydrogen bonding distances can be corrected by using the function d(O-H)=exp((2.173-d(O…O))/0.138)+0.958 obtained by least- squares fit to the data from the neutron diffraction at low temperatures. The valence-least-squares method is effective for the distance correction of very short hydrogen bond(<2.5 $\AA$). The distance correction is necessary for the long intermolecular hydrogen bond obtained from X-ray diffraction analysis.

• The Journal of Korean Institute of Electromagnetic Engineering and Science
• /
• v.22 no.6
• /
• pp.598-604
• /
• 2011

A method of miniaturizing branch line coupler with connected coupled lines is presented. The quarter-wavelength transmission lines in the typical microstrip branch line coupler are replaced with the connected coupled lines with same characteristics of ones for compact size. The connected coupled line is analyzed by T-equivalent circuit and Z parameters based on the even-odd mode analysis. The proposed branch line couplers with connected coupled line are designed and fabricated on FR4 substrate at 2.4 GHz. The measured results show good agreement with theoretical prediction. And the experimental results show that the size of coupler is 37 precent smaller than conventional coupler. This minimized coupler is suitable for Butler Matrix as feeder for mobile communication beam forming antenna.

• KIISE Transactions on Computing Practices
• /
• v.23 no.12
• /
• pp.698-702
• /
• 2017

Since the introduction of Deep Neural Networks to the Natural Language Processing field, two major approaches have been considered for modeling text. One method involved learning embeddings, i.e. the distributed representations containing abstract semantics of words or sentences, with the textual context. The other strategy consisted of composing the embeddings trained by the above to get embeddings of longer texts. However, most studies of the composition methods just adopt word embeddings without consideration of the optimal embedding unit and the optimal method of composition. In this paper, we conducted experiments to analyze the optimal embedding unit and the optimal composition method for modeling longer texts, such as documents. In addition, we suggest a new discourse-based composition to overcome the limitation of the sequential composition method on composing sentence embeddings.

• The Korean Journal of Applied Statistics
• /
• v.1 no.2
• /
• pp.10-26
• /
• 1987

If the forecasts from different, sources are combined in some way, the resulting forecasts may be more accurate than any of the individual components. In this paper, the established procedures of combining forecasts are reviewed and the alternative procedures are suggested. By the results of empirical analysis from survey data, the method of combining forecasts using the restricted regression weights, the restricted robust regression weights, and mixed regression weights are robust. We can not find the most efficient combined forecasts in any case if we select the corresponding decision by preliminary analysis for the statistical properties of individual dorecasts, our results of combined forecast can became useful.

• The Korean Journal of Applied Statistics
• /
• v.28 no.6
• /
• pp.1235-1243
• /
• 2015

Increased data volume in the ICT area has increased the importance of forecasting accuracy for internet traffic. Forecasting results may have paper plans for traffic management and control. In this paper, we propose combined forecasts based on several time series models such as Seasonal ARIMA and Taylor's adjusted Holt-Winters and Fractional ARIMA(FARIMA). In combined forecasting methods, we use simple-combined method, MSE based method (Armstrong, 2001), Ordinary Least Squares (OLS) method and Equality Restricted Least Squares (ERLS) method. The results show that the Seasonal ARIMA model outperforms in 3 hours ahead forecasts and that combined forecasts outperform in longer periods.

• The Korean Journal of Applied Statistics
• /
• v.21 no.6
• /
• pp.981-996
• /
• 2008

Statistical matching techniques whose aim is to achieve a complete data file from different sources. Since the statistical matching method proposed by Rubin (1986) assumes the multivariate normality for data, using this method to data which violates the assumption would involve some problems. This research proposed the statistical matching method using robust regression as an alternative to the linear regression. Furthermore, we carried out a simulation study to compare the performance of the robust regression model and the linear regression model for the statistical matching.

• Korean Journal of Optics and Photonics
• /
• v.7 no.3
• /
• pp.207-212
• /
• 1996

We propose an extended additional layer method (EALM) of calculating the coupling coefficient of arbitrary shaped diffraction gratings. In EALM, to determine the unperturbed field distribtution, a grating region is replaced by a new uniform layer whose dielectric constant is the average value of the dielectric constant of a grating region in both longitudinal and transverse directions. Using this method, we calculate the coupling coefficient for a five-layer distributed feedback structure device with trapezoidal and triangular gratings. The validity of this method is established by comparing the results calculated by partitioning the grating region up to five uniform layers.