• Journal of Digital Convergence
• /
• v.19 no.6
• /
• pp.251-258
• /
• 2021

In this paper, we proposed an encoder-decoder model utilizing residual learning to improve the accuracy of the U-Net-based semantic segmentation method. U-Net is a deep learning-based semantic segmentation method and is mainly used in applications such as autonomous vehicles and medical image analysis. The conventional U-Net occurs loss in feature compression process due to the shallow structure of the encoder. The loss of features causes a lack of context information necessary for classifying objects and has a problem of reducing segmentation accuracy. To improve this, The proposed method efficiently extracted context information through an encoder using residual learning, which is effective in preventing feature loss and gradient vanishing problems in the conventional U-Net. Furthermore, we reduced down-sampling operations in the encoder to reduce the loss of spatial information included in the feature maps. The proposed method showed an improved segmentation result of about 12% compared to the conventional U-Net in the Cityscapes dataset experiment.

• The Korean Journal of Applied Statistics
• /
• v.36 no.2
• /
• pp.129-139
• /
• 2023

Forecasting volatility is essential to avoiding the risk caused by the uncertainties of an financial asset. Complicated financial volatility features such as ambiguity between non-stationarity and stationarity, asymmetry, long-memory, sudden fairly large values like outliers bring great challenges to volatility forecasts. In order to address such complicated features implicity, we consider machine leaning models such as LSTM (1997) and GRU (2014), which are known to be suitable for existing time series forecasting. However, there are the problems of vanishing gradients, of enormous amount of computation, and of a huge memory. To solve these problems, a causal temporal convolutional network (TCN) model, an advanced form of 1D CNN, is also applied. It is confirmed that the overall forecasting power of TCN model is higher than that of the RNN models in forecasting VIX, VXD, and VXN, the daily volatility indices of S&P 500, DJIA, Nasdaq, respectively.

• The Bulletin of The Korean Astronomical Society
• /
• v.46 no.2
• /
• pp.46.3-47
• /
• 2021

Recently a multilayer spectral inversion (MLSI) model has been proposed to infer the physical parameters of plasmas in the solar chromosphere. The inversion solves a three-layer radiative transfer model using the strong absorption line profiles, H alpha and Ca II 8542 Å, taken by the Fast Imaging Solar Spectrograph (FISS). The model successfully provides the physical plasma parameters, such as source functions, Doppler velocities, and Doppler widths in the layers of the photosphere to the chromosphere. However, it is quite expensive to apply the MLSI to a huge number of line profiles. For example, the calculating time is an hour to several hours depending on the size of the scan raster. We apply deep neural network (DNN) to the inversion code to reduce the cost of calculating the physical parameters. We train the models using pairs of absorption line profiles from FISS and their 13 physical parameters (source functions, Doppler velocities, Doppler widths in the chromosphere, and the pre-determined parameters for the photosphere) calculated from the spectral inversion code for 49 scan rasters (~2,000,000 dataset) including quiet and active regions. We use fully connected dense layers for training the model. In addition, we utilize a skip connection to avoid a problem of vanishing gradients. We evaluate the model by comparing the pairs of absorption line profiles and their inverted physical parameters from other quiet and active regions. Our result shows that the deep learning model successfully reproduces physical parameter maps of a scan raster observation per second within 15% of mean absolute percentage error and the mean squared error of 0.3 to 0.003 depending on the parameters. Taking this advantage of high performance of the deep learning model, we plan to provide the physical parameter maps from the FISS observations to understand the chromospheric plasma conditions in various solar features.

• Journal of the Korea Society of Computer and Information
• /
• v.28 no.7
• /
• pp.11-18
• /
• 2023

In this paper, we evaluate deep learning time series forecasting models. Recent studies show that those models perform better than the traditional prediction model such as ARIMA. Among them, recurrent neural networks to store previous information in the hidden layer are one of the prediction models. In order to solve the gradient vanishing problem in the network, LSTM is used with small memory inside the recurrent neural network along with BI-LSTM in which the hidden layer is added in the reverse direction of the data flow. In this paper, we compared the performance of Informer by comparing with other models (LSTM, BI-LSTM, and Transformer) for real Nitrogen dioxide (NO2) data. In order to evaluate the accuracy of each method, mean square root error and mean absolute error between the real value and the predicted value were obtained. Consequently, Informer has improved prediction accuracy compared with other methods.

• Journal of the Chungcheong Mathematical Society
• /
• v.37 no.1
• /
• pp.27-40
• /
• 2024

In this paper, we consider the higher-order HartreeFock equations. The higher-order linear Schrödinger equation was introduced in [5] as the formal finite Taylor expansion of the pseudorelativistic linear Schrödinger equation. In [13], the authors established global-in-time Strichartz estimates for the linear higher-order equations which hold uniformly in the speed of light c ≥ 1 and as their applications they proved the convergence of higher-order Hartree-Fock equations to the corresponding pseudo-relativistic equation on arbitrary time interval as c goes to infinity when the Taylor expansion order is odd. To achieve this, they not only showed the existence of solutions in L² space but also proved that the solutions stay bounded uniformly in c. We address the remaining question on the convergence of higherorder Hartree-Fock equations when the Taylor expansion order is even. The distinguished feature from the odd case is that the group velocity of phase function would be vanishing when the size of frequency is comparable to c. Owing to this property, the kinetic energy of solutions is not coercive and only weaker Strichartz estimates compared to the odd case were obtained in [13]. Thus, we only manage to establish the existence of local solutions in H^s space for s > $\frac{1}{3}$ on a finite time interval [-T, T], however, the time interval does not depend on c and the solutions are bounded uniformly in c. In addition, we provide the convergence result of higher-order Hartree-Fock equations to the pseudo-relativistic equation with the same convergence rate as the odd case, which holds on [-T, T].

• Journal of Korean Society of Disaster and Security
• /
• v.12 no.2
• /
• pp.73-82
• /
• 2019

Recently, as the occurrence frequency of sudden floods due to climate change increased, the flood damage on riverside social infrastructures was extended so that there has been a threat of overflow. Therefore, a rapid prediction of potential flooding in riverside social infrastructure is necessary for administrators. However, most current flood forecasting models including hydraulic model have limitations which are the high accuracy of numerical results but longer simulation time. To alleviate such limitation, data driven models using artificial neural network have been widely used. However, there is a limitation that the existing models can not consider the time-series parameters. In this study the water surface elevation of the Hangang River bridge was predicted using the NARX model considering the time-series parameter. And the results of the ANN and RNN models are compared with the NARX model to determine the suitability of NARX model. Using the 10-year hydrological data from 2009 to 2018, 70% of the hydrological data were used for learning and 15% was used for testing and evaluation respectively. As a result of predicting the water surface elevation after 3 hours from the Hangang River bridge in 2018, the ANN, RNN and NARX models for RMSE were 0.20 m, 0.11 m, and 0.09 m, respectively, and 0.12 m, 0.06 m, and 0.05 m for MAE, and 1.56 m, 0.55 m and 0.10 m for peak errors respectively. By analyzing the error of the prediction results considering the time-series parameters, the NARX model is most suitable for predicting water surface elevation. This is because the NARX model can learn the trend of the time series data and also can derive the accurate prediction value even in the high water surface elevation prediction by using the hyperbolic tangent and Rectified Linear Unit function as an activation function. However, the NARX model has a limit to generate a vanishing gradient as the sequence length becomes longer. In the future, the accuracy of the water surface elevation prediction will be examined by using the LSTM model.

• Journal of the Korean Institute of Landscape Architecture
• /
• v.34 no.2 s.115
• /
• pp.80-98
• /
• 2006

This study originated from following questions. What can we understand the conception of deconstruction, which has been the core idea of new discourses developed in various ways since modernism? How can this question be interpreted in landscape design? What is the conceptional frame of integration the prominent hybrid post-genre movements and phenomena? The frame can be epitomized with the deconstruction phenomenon. 'Deconstruction' is the core conception appeared in late or post-modern ages in the embodiment of modernity and can be viewed as an integrating or a hybrid phenomenon between areas or genres in formative arts. Therefore, the author regards the hybrid movements widely witnessed in the post contemporary formative arts as one of the most important indicators of de-constructive signs. It is safe to say that the phenomenon of this integration or hybridism, of course, does not threaten the identity of landscape design but serves as an opportunity to extend the areas of landscape design. One of the consequences of this integration or hybridism is the voluntary participation of users who have been alienated in the production of the meanings of design works and hybrid landscape design with the hybridization of genres that is characterized with transformation in forms. This view is based on the distinction between hybridization of interactions between the designer (the subject) and the user (the object), and hybridization of synesthesia. Generally speaking, this is an act of destroying boundaries of the daily life and arts. At the same time, it corresponds to vanishing of modern aesthetics and emerging of post-contemporary aesthetics which is a new aesthetic category like sublimeness. This types of landscape design tries to restore humans' sensibility and perceptions restrained by rationality and recognition in previous approach and to express non-materialistic characteristics with precaution against excessive materialism in the modern era. In light of these backgrounds, the study aims to suggest the hybrid concept and to explorer a new landscape design approach with this concept, in order to change the design structure from 'completed' or 'closed' toward 'opened' and to understand the characteristics of interactions between users and designs. This new approach is expected to create an open-space integrating complexity and dynamics of users. At the same time, it emphasizes senses of user' body with synesthesia and non-determination. The focus is placed on user participation and sublimity rather than on aesthetic beauty, which kind of experience is called simulacre. By attaching importance to user participation, the work got free from the material characteristics, and acceptance from the old practice of simple perception and contemplation. The boundaries between the subject and object and the beautiful and ordinary, from the perspective of this approach, are vanished. Now everything ordinary can become an artistic work. Western dichotomy and discrimination is not effective any more. And there is 'de-construction' where there is perfect equality between ordinary daily life and beautiful arts. Thus today's landscape design pays attention to the user and uses newly perceived sensitivity by pursing obscure and unfamiliar things rather than aesthetic beauty. Space is accordingly defined to take place accidentally as happening and event, not as volume of shape. It's the true way to express spatiality of landscape design. That's an attempt to reject conventional concepts about forms and space, which served as the basis for landscape design, and to search for new things.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.37 no.6
• /
• pp.25-37
• /
• 2000

The trajectories of feature points can be defined by the correspondences between points in consecutive frames. The correspondence problem is known to be difficult to solve because false positives and false negatives almost always exist in real image sequences. In this paper, we propose a robust feature tracking algorithm considering incomplete trajectories such as entering and/or vanishing trajectories. The trajectories of feature points are determined by calculating the matching measure, which is defined as the minimum weighted Euclidean distance between two feature points. The weights are automatically updated in order to properly reflect the motion characteristics. We solve the correspondence problem as an optimal graph search problem, considering that the existence of false feature points may have serious effect on the correspondence search. The proposed algorithm finds a local optimal correspondence so that the effect of false feature point can be minimized in the decision process. The time complexity of the proposed graph search algorithm is given by O(mn) in the best case and O($m^2n$) in the worst case, where m and n arc the number of feature points in two consecutive frames. By considering false feature points and by properly reflecting motion characteristics, the proposed algorithm can find trajectories correctly and robustly, which has been shown by experimental results.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.18 no.10
• /
• pp.75-80
• /
• 2017

In this paper, we describe the effect of the regularization method and the network with identity mapping on the performance of the activation functions in deep convolutional neural networks. The activation functions act as nonlinear transformation. In early convolutional neural networks, a sigmoid function was used. To overcome the problem of the existing activation functions such as gradient vanishing, various activation functions were developed such as ReLU, Leaky ReLU, parametric ReLU, and ELU. To solve the overfitting problem, regularization methods such as dropout and batch normalization were developed on the sidelines of the activation functions. Additionally, data augmentation is usually applied to deep learning to avoid overfitting. The activation functions mentioned above have different characteristics, but the new regularization method and the network with identity mapping were validated only using ReLU. Therefore, we have experimentally shown the effect of the regularization method and the network with identity mapping on the performance of the activation functions. Through this analysis, we have presented the tendency of the performance of activation functions according to regularization and identity mapping. These results will reduce the number of training trials to find the best activation function.

• The Korean Journal of Physiology
• /
• v.2 no.2
• /
• pp.93-99
• /
• 1968

The Present study attempted to analyze the fate of CO diffused into the circulating blood through the alveoli. Dogs were induced to CO poisoning by rebreathing CO gas mixture contained in Krog's spirometer, by closed circuit method, for 60 minutes. The spirometer was filled initially with 282 ml of CO and 20 liters of air and oxygen, so the composition of gases were arranged as 1.4% in CO and 50% in $O_2$ at the begining of the rebreathing. Oxygen was added corresponding to the utilization of $O_2$ by the animal in proceeding of the experiment. At 60th minutes of CO rebreathing, the concentration of CO in arterial blood and in mixed venous blood were analysed and compared with each other after the CO contents were corrected with the hematocrit measured in the arterial and mixed venous blood. The distribution of CO gas to other tissues was estimated by the analysis of CO diffused into the cystic bile and into the peritoneal gas pocket which was formed by injection of 300 ml air into the peritoneal cavity prior to the CO gas rebreathing. The blood volume was measured by dilution method using $^{51}Chromium$ tagged red cells. CO amount vanished in the animal body was calculated by subtraction of total CO content in blood stream and the CO remained in closed circuit breathing system from the CO amount given to the breathing system at the begining of the experiment. Results obtained are summarized as follows: 1. The content of CO corrected by the hematocrit value was slightly less in mixed venous blood than in arterial blood. The amount of CO diffused into the cystic bile and into the peritoneal cavity was averaged to 0.1% and 0.4% of the CO amount in 100 ml of blood, respectively. 2. For 60 minutes of CO rebreathing, CO-hemoglobin saturation reached about 77% at the 60th minutes, CO amount vanished in the experimental animal averaged 36.1 ml/dog/hr., or 21% of the total CO volume in the blood stream. The average vanishing rate of CO during 60 minutes of CO rebreathing per kg of body weight was 2.71 ml/hr. Production of CO measured in ten dogs under hypoxic condition averaged 0.023 ml/kg/hr. The major part of the CO vanished in the dogs seemed to be oxidized to $CO_2$ by various tissues of the animal. The conclusion might be delivered as such oxidation of CO to $CO_2$ by animal tissues can play a role in part of the process of recovery and protection of animal from CO-poisoning.