• Proceedings of the Korea Water Resources Association Conference
• /
• 2015.05a
• /
• pp.225-225
• /
• 2015

Drought events usually evolve slowly in time and their impacts generally span a long period of time. This indicates that the sequence of drought is not completely random. The Hidden Markov Model (HMM) is a probabilistic model used to represent dependences between invisible hidden states which finally result in observations. Drought characteristics are dependent on the underlying generating mechanism, which can be well modelled by the HMM. This study employed a HMM with Gaussian emissions to fit the Standardized Precipitation Index (SPI) series and make multi-step prediction to check the drought characteristics in the future. To estimate the parameters of the HMM, we employed a Bayesian model computed via Markov Chain Monte Carlo (MCMC). Since the true number of hidden states is unknown, we fit the model with varying number of hidden states and used reversible jump to allow for transdimensional moves between models with different numbers of states. We applied the HMM to several stations SPI data in South Korea. The monthly SPI data from January 1973 to December 2012 was divided into two parts, the first 30-year SPI data (January 1973 to December 2002) was used for model calibration and the last 10-year SPI data (January 2003 to December 2012) for model validation. All the SPI data was preprocessed through the wavelet denoising and applied as the visible output in the HMM. Different lead time (T= 1, 3, 6, 12 months) forecasting performances were compared with conventional forecasting techniques (e.g., ANN and ARMA). Based on statistical evaluation performance, the HMM exhibited significant preferable results compared to conventional models with much larger forecasting skill score (about 0.3-0.6) and lower Root Mean Square Error (RMSE) values (about 0.5-0.9).

• Smart Structures and Systems
• /
• v.23 no.6
• /
• pp.589-613
• /
• 2019

Cables are prone to vibration due to their low inherent damping characteristics. Recently, negative stiffness dampers have gained attentions, because of their promising energy dissipation ability. The viscous inertial mass damper (termed as VIMD hereinafter) can be viewed as one realization of the inerter. It is formed by paralleling an inertial mass part with a common energy dissipation element (e.g., viscous element) and able to provide pseudo-negative stiffness properties to flexible systems such as cables. A previous study examined the potential of IMD to enhance the damping of stay cables. Because there are already models for common energy dissipation elements, the key to establish a general model for IMD is to propose an analytical model of the rotary mass component. In this paper, the characteristics of the rotary mass and the proposed analytical model have been evaluated by the numerical and experimental tests. First, a series of harmonic tests are conducted to show the performance and properties of the IMD only having the rotary mass. Then, the mechanism of nonlinearities is analyzed, and an analytical model is introduced and validated by comparing with the experimental data. Finally, a real-time hybrid simulation test is conducted with a physical IMD specimen and cable numerical substructure under distributed sinusoidal excitation. The results show that the chosen model of the rotary mass part can provide better estimation on the damper's performance, and it is better to use it to form a general analytical model of IMD. On the other hand, the simplified damper model is accurate for the preliminary simulation of the cable responses.

• The Korean Journal of Applied Statistics
• /
• v.34 no.1
• /
• pp.75-98
• /
• 2021

This paper compares three statistical models that examine the relationship between national and provincespecific fertility rates. The three models are two of the regression models and a cointegration model. The regression model is by substituting Gompit transformation for the cumulative fertility rate by the average for ten years, and this model applies the raw data without transformation of the fertility data. A cointegration model can be considered when fitting the unstable time series of fertility rate in probability process. This paper proposes the following when it is intended to derive the relation of non-stationary fertility rate between the national and provinces. The cointegrated relationship between national and regional fertility rates is first derived. Furthermore, if this relationship is not significant, it is proposed to look at the national and regional fertility rate relationships with a regression model approach using raw data without transformation. Also, the regression model method of substituting Gompit transformation data resulted in an overestimation of fertility rates compared to other methods. Finally, Seoul, Busan, Daegu, Incheon, Gwangju, Daejeon and Gyeonggi province are expected to show a total fertility rate of 1.0 or less from 2025 to 2030, so an urgent and efficient policy to raise this level is needed.

• Geomechanics and Engineering
• /
• v.27 no.5
• /
• pp.527-535
• /
• 2021

The creep characteristics of rock is of great significance for the study of long-term stability of engineering, so it is necessary to carry out indoor creep test and creep model of rock. First of all, in different water-bearing state and different positive pressure conditions, the granite is graded loaded to conduct indoor shear creep test. Through the test, the shear creep characteristics of granite are obtained. According to the test results, the stress-strain isochronous curve is obtained, and then the long-term strength of granite under different conditions is determined. Then, the fractional-order calculus software element is introduced, and it is connected in series with the spring element and the nonlinear viscoplastic body considering the creep acceleration start time to form a nonlinear viscoplastic creep model with fewer elements and fewer parameters. Finally, based on the shear creep test data of granite, using the nonlinear curve fitting of Origin software and Levenberg-Marquardt optimization algorithm, the parameter fitting and comparative analysis of the nonlinear creep model are carried out. The results show that the test data and the model curve have a high degree of fitting, which further explains the rationality and applicability of the established nonlinear visco-elastoplastic creep model. The research in this paper can provide certain reference significance and reference value for the study of nonlinear creep model of rock in the future.

• Journal of the Korea Society of Computer and Information
• /
• v.27 no.8
• /
• pp.61-68
• /
• 2022

In this paper, we propose a system to predict the GPS trajectory of a pedestrian based on a deep learning model. Pedestrian trajectory prediction is a study that can prevent pedestrian danger and collision situations through notifications, and has an impact on business such as various marketing. In addition, it can be used not only for pedestrians but also for path prediction of unmanned transportation, which is receiving a lot of spotlight. Among various trajectory prediction methods, this paper is a study of trajectory prediction using GPS data. It is a deep learning model-based study that predicts the next route by learning the GPS trajectory of pedestrians, which is time series data. In this paper, we presented a data set construction method that allows the deep learning model to learn the GPS route of pedestrians, and proposes a trajectory prediction deep learning model that does not have large restrictions on the prediction range. The parameters suitable for the trajectory prediction deep learning model of this study are presented, and the model's test performance are presented.

• Journal of The Korean Society of Agricultural Engineers
• /
• v.64 no.6
• /
• pp.35-41
• /
• 2022

Heavy snow damage can be prevented in advance with an appropriate security system. To develop the security system, we developed a model that predicts snow depth after a few hours when the snow depth is observed, and utilized it to calculate a failure probability with various types of greenhouses and observed snow depth data. We compared the Markov chain model and Bayesian long short-term memory models with varying input data. Markov chain model showed the worst performance, and the models that used only past snow depth data outperformed the models that used other weather data with snow depth (temperature, humidity, wind speed). Also, the models that utilized 1-hour past data outperformed the models that utilized 3-hour data and 6-hour data. Finally, the Bayesian LSTM model that uses 1-hour snow depth data was selected to predict snow depth. We compared the selected model and the shifting method, which uses present data as future data without prediction, and the model outperformed the shifting method when predicting data after 11-24 hours.

• Journal of the Korea Society of Computer and Information
• /
• v.27 no.5
• /
• pp.29-36
• /
• 2022

In this paper, we propose an efficient agricultural products price prediction model based on dataset which provided in DACON. This model is XGBoost and CatBoost, and as an algorithm of the Gradient Boosting series, the average accuracy and execution time are superior to the existing Logistic Regression and Random Forest. Based on these advantages, we design a machine learning model that predicts prices 1 week, 2 weeks, and 4 weeks from the previous prices of agricultural products. The XGBoost model can derive the best performance by adjusting hyperparameters using the XGBoost Regressor library, which is a regression model. The implemented model is verified using the API provided by DACON, and performance evaluation is performed for each model. Because XGBoost conducts its own overfitting regulation, it derives excellent performance despite a small dataset, but it was found that the performance was lower than LGBM in terms of temporal performance such as learning time and prediction time.

• Journal of Korea Water Resources Association
• /
• v.33 no.6
• /
• pp.783-791
• /
• 2000

A multidimensional representation for precipitation, given In the theory proposed by E. Waymire et al. (1984), is used for simulating rainfall in space and time. The model produces moving storms with realistic meso-scale meteorological features in time and space. The first- and second-order statistics derived from observed JX)int gauge data were used to estimate the model parameters based on the Nelder-Mead algorithm of optimization. Then twelve-year traces of rainfall intensities at fixed gage stations were generated at intervals of 1 hours. First- and second-order statistics are evaluated from the above series, which are used for estimating the parameters of one dimensional model of temporal rainfall at a point. As a result from the comparisons of one dimensional model parameters used observed and generated data from multidimensional model, we found that the multidimensional rainfall model generated visually realistic spatial patterns of rainfall as well as realistic temporal hyetographs of rainfall at a point. point.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.5
• /
• pp.1431-1445
• /
• 2022

We construct a deep neural network model named ECGResNet. This model can diagnosis diseases based on 12-lead ECG data of eight common cardiovascular diseases with a high accuracy. We chose the 16 Blocks of ResNet50 as the main body of the model and added the Squeeze-and-Excitation module to learn the data information between channels adaptively. We modified the first convolutional layer of ResNet50 which has a convolutional kernel of 7 to a superposition of convolutional kernels of 8 and 16 as our feature extraction method. This way allows the model to focus on the overall trend of the ECG signal while also noticing subtle changes. The model further improves the accuracy of cardiovascular and cerebrovascular disease classification by using a fully connected layer that integrates factors such as gender and age. The ECGResNet model adds Dropout layers to both the residual block and SE module of ResNet50, further avoiding the phenomenon of model overfitting. The model was eventually trained using a five-fold cross-validation and Flooding training method, with an accuracy of 95% on the test set and an F1-score of 0.841.We design a new deep neural network, innovate a multi-scale feature extraction method, and apply the SE module to extract features of ECG data.

• Asia-Pacific Journal of Business
• /
• v.15 no.2
• /
• pp.207-222
• /
• 2024

Purpose - This study investigates whether the lagged price difference ratio between preferred and common stocks is related to the return and closing price of the preferred stock using three panel models. Design/methodology/approach - As a first step, we use a two-way fixed effect panel model with stationary preferred stock returns as a dependent variable. For robustness, we then apply the autoregressive distributed lag model (ARDL) and error correction model (ECM) with nonstationary closing prices of the preferred stocks as a dependent variable and compare the results of each model. The ARDL and ECM models provide an advantage of estimating a long-run equilibrium equation together if a long-run relationship exists between the two time-series variables compared to the fixed effect model. Findings - Our sample consists of 107 preferred stocks with at least four years of daily observations as of the end of December 2023. The coefficients of the error correction terms in the ARDL and ECM models are highly statistically significant, approximately -0.08. This indicates that the disequilibrium between the closing prices of common and preferred stocks adjusts by about 8% per day toward equilibrium. In all three models, the price difference ratio on day t-1 was statistically significant in explaining the preferred stock returns or closing prices on day t, implying that trading based on the previous day's price difference ratio is effective for one day. Research implications or Originality - Furthermore, the returns on preferred stocks are higher for firms with a lower proportion of foreign investors or a lower foreign market capitalization of preferred stocks. This suggests that foreign investors with informational advantages do not actively engage in profit-taking by trading preferred stocks, thus not narrowing the price difference. In summary, the recent surge in preferred stock prices is likely driven mainly by the irrational behavior of retail investors.