• Journal of Ship and Ocean Technology
• /
• v.5 no.1
• /
• pp.38-54
• /
• 2001

A finite difference method for calculating turbulent flow and surface wave fields around a ship model is evaluated through the comparison with the experimental data of a Series 60 $C_B$=0.6 ship model. The method solves the Reynolds-averaged Navior-Stokes Equations using the non-staggered grid system, the four-stage Runge-Kutta scheme for the temporal integration of governing equations and the Bladwin-Lomax model for the turbulence closure. The free surface waves are captured by solving the equation of the kinematic free-surface condition using the Lax-Wendroff scheme and free-surface conforming grids are generated at each time step so that one of the grid surfaces coincides always with the free surface. The computational results show an overall close agreement with the experimental data and verify that the present method can simulate well the turbulent boundary layers and wakes as well as the free-surface waves.

• Journal of the Institute of Electronics Engineers of Korea TE
• /
• v.39 no.1
• /
• pp.83-90
• /
• 2002

This paper proposed the strategy learning method by means of the fusion of Back-Propagation neural network and Q learning algorithm for two-person, deterministic janggi board game. The learning process is accomplished simply through the playing each other. The system consists of two parts of move generator and search kernel. The one consists of move generator generating the moves on the board, the other consists of back-propagation and Q learning plus $\alpha$$\beta$ search algorithm in an attempt to learn the evaluation function. while temporal difference learns the discrepancy between the adjacent rewards, Q learning acquires the optimal policies even when there is no prior knowledge of effects of its moves on the environment through the learning of the evaluation function for the augmented rewards. Depended on the evaluation function through lots of games through the learning procedure it proved that the percentage won is linearly proportional to the portion of learning in general.

• International Journal of Vascular Biomedical Engineering
• /
• v.1 no.1
• /
• pp.13-23
• /
• 2003

Backgrounds: The present study in angulated coronary stenosis was to evaluate the influence of velocity and wall shear stress (WSS) on coronary atherosclerosis, the changes of hemodynamic indices following coronary stenting, as well as their effect of evolving in-stent restenosis using human in vivo hemodynamic parameters and computed simulation quantitatively and qualitatively. Methods: Initial and follow-up coronary angiographies in the patients with angulated coronary stenosis were performed (n=80). Optimal coronary stenting in angulated coronary stenosis had two models: < 50 % angle changed(model 1, n=43), > 50% angle changed group (model 2, n=37) according to percent change of vascular angle between pre- and post-intracoronary stenting. Flow-velocity wave obtained from in vivo intracoronary Doppler study data was used for in vitro numerical simulation. Spatial and temporal patterns of velocity vector and recirculation area were drawn throughout the selected segment of coronary models. WSS of pre/post-intracoronary stenting were calculated from three-dimensional computer simulation. Results: Follow-up coronary angiogram demonstrated significant difference in the percent of diameter stenosis between two groups (group 1: $40.3{\pm}30.2$ vs. group 2: $25.5{\pm}22.5%$, p<0.05). Negative WSS area on 3D simulation, which is consistent with re-circulation area of velocity vector, was noted on the inner wall of post-stenotic area before stenting. The negative WSS was disappeared after stenting. High spatial and temporal WSS before stenting fell into within physiologic WSS after stenting. This finding was prominent in Model 2 (p<0.01) Conclusions: The present study suggests that hemodynamic forces exerted by pulsatile coronary circulation termed as WSS might affect on the evolution of atherosclerosis within the angulated vascular curvature. Moreover, geometric change, such as angular difference between pre / post-intracoronary stenting might give proper information of optimal hemodynamic charateristics for vascular repair after stenting.

• KOREAN JOURNAL OF CROP SCIENCE
• /
• v.50 no.4
• /
• pp.268-275
• /
• 2005

Early predictions of crop yields call provide information to producers to take advantages of opportunities into market places, to assess national food security, and to provide early food shortage warning. The objectives of this study were to identify the most useful parameters for estimating yields and to compare two model selection methods for finding the 'best' model developed by multiple linear regression. This research was conducted in two 65ha corn/soybean rotation fields located in east central South Dakota. Data used to develop models were small temporal variability information (STVI: elevation, apparent electrical conductivity $(EC_a)$, slope), large temporal variability information (LTVI : inorganic N, Olsen P, soil moisture), and remote sensing information (green, red, and NIR bands and normalized difference vegetation index (NDVI), green normalized difference vegetation index (GDVI)). Second order Akaike's Information Criterion (AICc) and Stepwise multiple regression were used to develop the best-fitting equations in each system (information groups). The models with $\Delta_i\leq2$ were selected and 22 and 37 models were selected at Moody and Brookings, respectively. Based on the results, the most useful variables to estimate corn yield were different in each field. Elevation and $EC_a$ were consistently the most useful variables in both fields and most of the systems. Model selection was different in each field. Different number of variables were selected in different fields. These results might be contributed to different landscapes and management histories of the study fields. The most common variables selected by AICc and Stepwise were different. In validation, Stepwise was slightly better than AICc at Moody and at Brookings AICc was slightly better than Stepwise. Results suggest that the Alec approach can be used to identify the most useful information and select the 'best' yield models for production fields.

• Journal of KIISE:Software and Applications
• /
• v.29 no.11
• /
• pp.785-794
• /
• 2002

Automatic stock trading systems should be able to solve various kinds of optimization problems such as market trend prediction, stock selection, and trading strategies, in a unified framework. But most of the previous trading systems based on supervised learning have a limit in the ultimate performance, because they are not mainly concerned in the integration of those subproblems. This paper proposes a stock trading system, called R-Trader, based on reinforcement teaming, regarding the process of stock price changes as Markov decision process (MDP). Reinforcement learning is suitable for Joint optimization of predictions and trading strategies. R-Trader adopts two popular reinforcement learning algorithms, temporal-difference (TD) and Q, for selecting stocks and optimizing other trading parameters respectively. Technical analysis is also adopted to devise the input features of the system and value functions are approximated by feedforward neural networks. Experimental results on the Korea stock market show that the proposed system outperforms the market average and also a simple trading system trained by supervised learning both in profit and risk management.

• Journal of Korea Water Resources Association
• /
• v.47 no.4
• /
• pp.331-342
• /
• 2014

In water resources planning, to decide proper water supply capacity is a very important task. Once water supply system such as a dam is decided, it will affect whole range of water resources circumstances for a long time. Even though systematic approaches have been implemented since 1980, many problems are still prevail in reality. Especially some issues related to the reliability analysis method used in planning dams in Korea have been persistently brought up. This study is to diagnose problems on the reliability criterion in water supply capacity assessment of water resources systems and discuss a valid method. As a result, the estimates by the different analysis time intervals, in case of the temporal reliability, show no large difference, but there is a large difference when assessment time intervals are differently applied. The volumetric reliability accounts for 2~3% higher than that of the temporal reliability, and resiliency and vulnerability also show large differences by the analysis time intervals.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.2 no.3
• /
• pp.46-59
• /
• 1999

Remote sensing provides informations on the changes of the hydrological states and variables over with the temporal and spatial distribution to monitor hydrological conditions and changes for large area. Especially, it can extract a spatial distribution of hydrological parameters such as surface albedo, vegetation informations, and surface temperature to effectively manage water resources of the watershed. In this study, we analyzed the characteristic of temporal and spatial changes in surface hydrological parameters which is necessary to identify the spatial distribution of water resources. 5 Landsat TM data of 1995 which is collected for Bochong-chon watershed, located in the upper stream of Keum River, were used to estimate characteristics on the change of hydrological parameters and atmospheric correction was carried out using COST model. The study showed that the difference of the albedo by the land cover was very sensitive depending upon the change of sun elevation and the amount of water in the soil. The difference between the surface temperature analysis and the measured air temperature was from $2.5^{\circ}C$ to $3.86^{\circ}C$.

• The KIPS Transactions:PartB
• /
• v.10B no.6
• /
• pp.587-592
• /
• 2003

Many real world control problems have continuous states and actions. When the state space is continuous, the reinforcement learning problems involve very large state space and suffer from memory and time for learning all individual state-action values. These problems need function approximators that reason action about new state from previously experienced states. We introduce Fuzzy Q-Map that is a function approximators for 1 - step Q-learning and is based on fuzzy clustering. Fuzzy Q-Map groups similar states and chooses an action and refers Q value according to membership degree. The centroid and Q value of winner cluster is updated using membership degree and TD(Temporal Difference) error. We applied Fuzzy Q-Map to the mountain car problem and acquired accelerated learning speed.

• Korean Journal of Remote Sensing
• /
• v.26 no.2
• /
• pp.87-98
• /
• 2010

For generating accurate land cover map over the whole Korean Peninsula, post-mosaic classification method is desirable in large area where multiple image data sets are used. We try to derive an optimal mosaic method of multi-temporal Landsat ETM+ scenes for the land cover classification over the Korea Peninsula. Total 65 Landsat ETM+ scenes were acquired, which were taken in 2000 and 2001. To reduce radiometric difference between adjacent Landsat ETM+ scenes, we apply three relative radiometric correction methods (histogram matching, 1st-regression method referenced center image, and 1st-regression method at each Landsat ETM+ path). After the relative correction, we generated three mosaic images for three seasons of leaf-off, transplanting, leaf-on season. For comparison, three mosaic images were compared by the mean absolute difference and computer classification accuracy. The results show that the mosaic image using 1st-regression method at each path show the best correction results and highest classification accuracy. Additionally, the mosaic image acquired during leaf-on season show the higher radiance variance between adjacent images than other season.

• KIPS Transactions on Computer and Communication Systems
• /
• v.10 no.5
• /
• pp.155-162
• /
• 2021

n-step TD learning is a combination of Monte Carlo method and one-step TD learning. If appropriate n is selected, n-step TD learning is known as an algorithm that performs better than Monte Carlo method and 1-step TD learning, but it is difficult to select the best values of n. In order to solve the difficulty of selecting the values of n in n-step TD learning, in this paper, using the characteristic that overestimation of Q can improve the performance of initial learning and that all n-step returns have similar values for Q ≈ Q^*, we propose a new learning target, which is composed of the maximum and the mean of all k-step returns for 1 ≤ k ≤ n. Finally, in OpenAI Gym's Atari game environment, we compare the proposed algorithm with n-step TD learning and proved that the proposed algorithm is superior to n-step TD learning algorithm.