• Journal of Information Processing Systems
• /
• 제17권4호
• /
• pp.818-833
• /
• 2021

To address the problems of low precision rate, insufficient feature extraction, and poor contextual ability in existing text sentiment analysis methods, a mixed model account of a CNN-BiLSTM-TE (convolutional neural network, bidirectional long short-term memory, and topic extraction) model was proposed. First, Chinese text data was converted into vectors through the method of transfer learning by Word2Vec. Second, local features were extracted by the CNN model. Then, contextual information was extracted by the BiLSTM neural network and the emotional tendency was obtained using softmax. Finally, topics were extracted by the term frequency-inverse document frequency and K-means. Compared with the CNN, BiLSTM, and gate recurrent unit (GRU) models, the CNN-BiLSTM-TE model's F1-score was higher than other models by 0.0147, 0.006, and 0.0052, respectively. Then compared with CNN-LSTM, LSTM-CNN, and BiLSTM-CNN models, the F1-score was higher by 0.0071, 0.0038, and 0.0049, respectively. Experimental results showed that the CNN-BiLSTM-TE model can effectively improve various indicators in application. Lastly, performed scalability verification through a takeaway dataset, which has great value in practical applications.

• 한국광학회지
• /
• 제10권4호
• /
• pp.335-341
• /
• 1999

홀로그램 기록 매질로 acrylamide를 기초한 포토폴리머를 제작하고 광화학 표백 특성을 측정하였다. 포토폴리머의 광화학 표백 특성을 설명하기 위해 잘 알려진 Beer-Lambert 모델과 간단한 에너지 qosem 모델로부터 얻어진 율방정식에 기초한 광표백 모델에 관한 이론들을 유도하였다. 본 연구에 사용된 포토폴리머에 대해 Beer-Lambert 모델은 짧은 시간 영역에 대해서만 실험 결과와 일치하였으나 율방정식에 기초한 이론적인 광표백 모델은 짧고 긴 시간 영역 모두에 대해서 실험 결과와 잘 일치하였다.

• 한국환경과학회지
• /
• 제8권4호
• /
• pp.485-490
• /
• 1999

We will calculate concentration of air pollutants using ISCST3, FDM and AERMOD of models recommended in U. S. EPA which are able to predict concentration of short term for point source, complex like industrial complex, power plant and burn-up institution. Before executing model, as analyzing computational result of many cases according to selecting of input data, we will increasing predictable ability of model in limit range of model. Especially, we analyzed three cases-case of considering various emission rate according to time scale and not, case considering effect of atmospheric pollution materials removed by physical process. In our study, after comparing and analyzing results of three model, we choose the atmospheric dispersion model reflected well the characteristic of the area. And we will investigate how large the complex pollutant sources such as industrial complex contribute to atmospheric environment and air quality of the surrounding the area as predicting and estimating chosen model.

• Journal of Biosystems Engineering
• /
• 제12권2호
• /
• pp.38-43
• /
• 1987

An experimental study was conducted to develop a thin-layer rewetting equation of short grain rough rice of Akihikari variety. Four thin-layer rewetting equations were experimentally determined from $25^{\circ}C$ to $45^{\circ}C$ and 70%RH to 85%RH conditions. Diffusion, Henderson, Page, and Thompson equations widely used as thin-layer drying equations were selected. Experimental data were fitted to these equations using linear regression analysis except diffusion equation. The diffusivity in the diffusion equation was determined by optimization method. Four equations were highly significant. In order to compare the goodness of fit of each equation, the error mean square of each equawas calculated. The diffusion model was not a very good model because the error mean square was very large. The other three models showed the same level or error mean square and could predict satisfactorily the rewetting rate or short grain rough rice.

• 대한의용생체공학회:의공학회지
• /
• 제24권3호
• /
• pp.151-158
• /
• 2003

심장혈관 신호에 있어서 단기간의 beat-to-beat 변이(variability)에 대한 여러 연구에서 선형 분석기법들이 사용되었다. 그러나 단기간 beat-to-beat 변이에 대해 선형기법 사용의 타당성에 대한 연구나 선형과 비선형 특성을 비교한 연구는 수행되지 않았다. 본 논문의 목적은 단기간 beat-to-beat 변이의 비선형성 특성을 조사함으로써 선형기법 사용의 적절함을 증명하고자 한다. 이를 위해 선형 ARMA와 비선형 신경망(NN) 모델을 사용하여 예측을 수행하였는데, 과거의 순시 심박(HR)과 평균 혈압(BP)으로부터 현재의 심박과 혈압 예측을 상호 비교하였다. 이러한 예측모델을 평가하기 위해 MIMIC 데이터베이스로부터 HR와 BP 시계열을 사용하였다. 실험결과에 의하면 신경망에 의한 비선형성은 단기간 beat-to-beat 변이를 생성하는 시스템 동특성을 나타내는데 의미있는 역할을 하지 못하였으며, 이 사실은 ARMA 선형 분석기법이 이러한 시스템 동특성을 나타내는데 적절함을 보여주고 있다

• Asian Journal of Business Environment
• /
• 제11권4호
• /
• pp.17-29
• /
• 2021

Purpose: The aim of this study is to explore the dynamic linkage between conditional stock market volatility and macroeconomic uncertainty of Bangladesh. Research design, data, and methodology: This study uses monthly data covering the time period from January 2005 to December 2018. A comprehensive set of macroeconomic variables, namely industrial production index (IP), consumer price index (CPI), broad money supply (M2), 91-day treasury bill rate (TB), treasury bond yield (GB), exchange rate (EX), inflow of foreign remittance (RT) and stock market index of DSEX are used for analysis. Symmetric and asymmetric univariate GARCH family of models and multivariate VAR model, along with block exogeneity and impulse response functions, are implemented on conditional volatility series to discover the possible interactions and causal relations between macroeconomic forces and stock return. Results: The analysis of the study exhibits time-varying volatility and volatility persistence in all the variables of interest. Moreover, the asymmetric effect is found significant in the stock return and most of the growth series of macroeconomic fundamentals. Results from the multivariate VAR model indicate that only short-term interest rate significantly influence the stock market volatility, while conditional stock return volatility is significant in explaining the volatility of industrial production, inflation, and treasury bill rate. Conclusion: The findings suggest an increasing interdependence between the money market and equity market as well as the macroeconomic fundamentals of Bangladesh.

• The Journal of Asian Finance, Economics and Business
• /
• 제8권10호
• /
• pp.229-238
• /
• 2021

The article applies the ECM - ARDL model to examine the relationship between economic indicators and the existence of the disease in the long run of 10 ASEAN countries from 2000 to 2019. There are two models: The first model investigates the impact of GDP per capita, net inflow FDI, unemployment rate, and inflation rate on the proportion of export to GDP of ASEAN countries, the second model is similar to the first one but adds one more variable to the independent variable list - 'the variable for disease'. The results prove the long-run effect of GDP per capita, FDI, unemployment and inflation rate on export of the selected countries, though individual country shows differences in the sign and magnitude of these impacts. Surprisingly, the number of people suffering from disease does not affect the export of all selected countries as expected. The results of the two models also indicate that the disequilibrium in the short run converges to the equilibrium in the long run with a high proportion, especially in the case of Cambodia and the Philippines, with the rate of 95.65% and 151.94%, respectively. The findings can be useful for policymakers in promulgating efficient policies to enhance the trading activities of the selected countries.

• The Journal of Asian Finance, Economics and Business
• /
• 제8권11호
• /
• pp.47-66
• /
• 2021

Despite the dominance of the USD as a vehicle currency in non-US trade, most studies on the exchange rate-trade balance relationship ignore its importance. Some recent J-curve papers have proved that incorporating the role of USD as vehicle currency as a crucial determinant of trade balance can well reflect the reality of global trade and provide more detailed findings. Motivated by this new approach and by the fact that USD is substantially used in the trade between China and the EU and the UK, this paper scrutinizes how the vehicle currency USD and the bilateral exchange rates asymmetrically affect China's trade balance with each EU country and the UK. The results of NARDL estimation indicate that the USD models outperform the bilateral exchange rate (BER) models in terms of detecting significant long-run and short-run coefficients, which confirms the usefulness of the new approach. Also, this paper finds that the USD/CNY exchange rate cannot be neglected in China's trade with the EU and the UK, which can supplement China's policies on international trade and foreign exchange management.

• 한국경영과학회:학술대회논문집
• /
• 한국경영과학회 2003년도 추계학술대회 및 정기총회
• /
• pp.11-14
• /
• 2003

The dynamics of unobservable short rate are frequently estimated directly by using a proxy. We estimate the biases resulting from this practice ("proxy problem"). To solve this problem, State-Space models have been proposed by many researchers. State-Space models have been used to estimate the unobservable variables from the observable variables in econometrics. However, applications of State-Space models often result in a misleading interpretation of the underlying processes especially when the absorbability of the State-Space model and the assumption of noise processes in the state vector are not properly considered. In this study, we propose the exact State-Space model that properly considers the faults of previous researchers to solve the proxy problem.

• 재무관리연구
• /
• 제24권3호
• /
• pp.153-186
• /
• 2007

It is well known that the distributional properties of financial asset returns exhibit fatter-tails and skewer-mean than the assumption of normal distribution. The correct assumption of return distribution might improve the estimated performance of the Value-at-Risk(VaR) models in financial markets. In this paper, we estimate and compare the VaR performance using the RiskMetrics, GARCH and FIGARCH models based on the normal and skewed-Student-t distributions in two daily returns of the Korean Composite Stock Index(KOSPI) and Korean Won-US Dollar(KRW-USD) exchange rate. We also perform the expected shortfall to assess the size of expected loss in terms of the estimation of the empirical failure rate. From the results of empirical VaR analysis, it is found that the presence of long memory in the volatility of sample returns is not an important in estimating an accurate VaR performance. However, it is more important to consider a model with skewed-Student-t distribution innovation in determining better VaR. In short, the appropriate assumption of return distribution provides more accurate VaR models for the portfolio managers and investors.