• Proceedings of the Korea Water Resources Association Conference
• /
• 2019.05a
• /
• pp.363-363
• /
• 2019

활성화 함수(activation function)는 기계학습(machine learning)의 학습과정에 비선형성을 도입하여 심층적인 학습을 용이하게 하고 예측의 정확도를 높이는 중요한 요소 중 하나이다(Roy et al., 2019). 일반적으로 기계학습에서 사용되고 있는 활성화 함수의 종류에는 계단 함수(step function), 시그모이드 함수(sigmoid 함수), 쌍곡 탄젠트 함수(hyperbolic tangent function), ReLU 함수(Rectified Linear Unit function) 등이 있으며, 예측의 정확도 향상을 위하여 다양한 형태의 활성화 함수가 제시되고 있다. 본 연구에서는 기계학습을 통하여 수위예측 시 정확도 향상을 위하여 Hybrid 활성화 함수를 제안하였다. 연구대상지는 조수간만의 영향을 받는 한강을 대상으로 선정하였으며, 2009년 ~ 2018년까지 10년간의 수문자료를 활용하였다. 수위예측 알고리즘은 Python 내 Tensorflow의 RNN (Recurrent Neural Networks) 모델을 이용하였으며, 강수량, 수위, 조위, 댐 방류량, 하천 유량의 수문자료를 학습시켜 3시간 및 6시간 후의 수위를 예측하였다. 예측정확도 향상을 위하여 입력 데이터는 정규화(Normalization)를 시켰으며, 민감도 분석을 통하여 신경망모델의 은닉층 개수, 학습률의 최적 값을 도출하였다. Hybrid 활성화 함수는 쌍곡 탄젠트 함수와 ReLU 함수를 혼합한 형태로 각각의 가중치($w_1,w_2,w_1+w_2=1$)를 변경하여 정확도를 평가하였다. 그 결과 가중치의 비($w_1/w_2$)에 따라서 예측 결과의 RMSE(Roote Mean Square Error)가 최소가 되고 NSE (Nash-Sutcliffe model Efficiency coefficient)가 최대가 되는 지점과 Peak 수위의 예측정확도가 최대가 되는 지점을 확인할 수 있었다. 본 연구는 현재 Data modeling을 통한 수위예측의 정확도 향상을 위해 기초가 되는 연구이나, 향후 다양한 형태의 활성화 함수를 제안하여 정확도를 향상시킨다면 예측 결과를 통하여 침수예보에 대한 의사결정이 가능할 것으로 기대된다.

• Proceedings of the Korean Society for Bioinformatics Conference
• /
• 2003.10a
• /
• pp.26-51
• /
• 2003

Large scale protein interaction maps provide a new, global perspective with which to analyse protein function. PSIMAP, the Protein Structural Interactome Map, is a database of all the structurally observed interactions between superfamilies of protein domains with known three-dimensional structure in thePDB. PSIMAP incorporates both functional and evolutionary information into a single network. It makes it possible to age protein domains in terms of taxonomic diversity, interaction and function. One consequence of it is to predict the most important protein domain structure in evolution. We present a global analysis of PSIMAP using several distinct network measures relating to centrality, interactivity, fault-tolerance, and taxonomic diversity. We found the following results: ${\bullet}$ Centrality: we show that the center and barycenter of PSIMAP do not coincide, and that the superfamilies forming the barycenter relate to very general functions, while those constituting the center relate to enzymatic activity. ${\bullet}$ Interactivity: we identify the P-loop and immunoglobulin superfamilies as the most highly interactive. We successfully use connectivity and cluster index, which characterise the connectivity of a superfamily's neighbourhood, to discover superfamilies of complex I and II. This is particularly significant as the structure of complex I is not yet solved. ${\bullet}$ Taxonomic diversity: we found that highly interactive superfamilies are in general taxonomically very diverse and are thus amongst the oldest. This led to the prediction of the oldest and most important protein domain in evolution of lift. ${\bullet}$ Fault-tolerance: we found that the network is very robust as for the majority of superfamilies removal from the network will not break up the network. Overall, we can single out the P-loop containing nucleotide triphosphate hydrolases superfamily as it is the most highly connected and has the highest taxonomic diversity. In addition, this superfamily has the highest interaction rank, is the barycenter of the network (it has the shortest average path to every other superfamily in the network), and is an articulation vertex, whose removal will disconnect the network. More generally, we conclude that the graph-theoretic and taxonomic analysis of PSIMAP is an important step towards the understanding of protein function and could be an important tool for tracing the evolution of life at the molecular level.

• The Journal of the Korean bone and joint tumor society
• /
• v.3 no.1
• /
• pp.18-25
• /
• 1997

Prosthetic replacement is one of the most common methods of reconstruction after resection of malignant tumor around the knee. Gait analysis provides a relative objective data about the gait function of patients with prosthesis. The purpose of this study was to compare the gait pattern of the patients who underwent limb salvage surgery with prosthesis for distal femur and that of patients with prosthesis for proximal tibia. This study included ten patients (4 males, 6 females, mean age 22.7 years, range 14-36) who underwent a wide resection and Kotz hinged modular reconstruction prosthesis replacement and six normal adult(Control). The site of bone tumor was the distal femur (Group 1) in six patients and proximal tibia (Group 2) in 4 patients. The follow-up period ranged from 15 to 82 months (mean : 33 months). The evaluation consisted of clinical assessment, radiographic assessment, gait analysis using VICON 370 Motion Analysis System. The gait analysis included the linear parameters such as, walking velocity, cadence, step length, stride length, stance time, swing time, single support and double support time and the three-dimensional kinematics (joint rotation angle, velocity of joint rotation) of ankle, knee, hip and pelvis in sagittal, coronal and transverse plane. For the kinetic evaluation, the moment of force (unit: Nm/kg) and power (unit: Watt/kg) of ankle, knee and hip joint in sagittal, coronal and transverse plane. In the linear parameters, cadence, velocity, step time and single support were decreased in both group 1 and group 2 compared with control. Double support decreased in group 2 compared with control significantly(p<.05). In contrast to our hypothesis, there was no significant difference between group 1 and group 2. In Kinematics, we observed significant difference (p<.05) of decreased knee flexion in loading response (G2

• The Korean Journal of Applied Statistics
• /
• v.30 no.4
• /
• pp.579-590
• /
• 2017

The multivariate empirical distribution function could be defined when its distribution function can be estimated. It is known that bivariate empirical distribution functions could be visualized by using Step plot and Quantile plot. In this paper, the multivariate empirical distribution plot is proposed to represent the multivariate empirical distribution function on the unit square. Based on many kinds of empirical distribution plots corresponding to various multivariate normal distributions and other specific distributions, it is found that the empirical distribution plot also depends sensitively on its distribution function and correlation coefficients. Hence, we could suggest five goodness-of-fit test statistics. These critical values are obtained by Monte Carlo simulation. We explore that these critical values are not much different from those in text books. Therefore, we may conclude that the proposed test statistics in this work would be used with known critical values with ease.

• 제어로봇시스템학회:학술대회논문집
• /
• 2004.08a
• /
• pp.248-251
• /
• 2004

Stability of truck and trailer are the most significance in Thai automotive industry. This paper presents the mathematical model of a six-degree-of-freedom semi-trailer vehicle. Search method was implemented to obtain the optimum design variables of the trailer which are the distance from the fifth wheel to the centroid of the trailer and the distance from the centroid of the trailer to the trailer axel. The objective function is to minimize the steady side slip velocity, steady-state yawing velocity and steady-state angle between the tractor and the trailer. From the calculation , the optimum distance from the fifth wheel to the centroid of the trailer and the optimum distance from the centroid of the trailer to the trailer axle are 5.50 and 3.25 meters respectively. The stability of the optimal semi-trailer vehicle was also examined in steady state. The steady side slip velocity, yawing velocity and the angle between tractor and trailer are also obtained using linearization technique under unit step disturbance of the tractor front wheel steering angle.

• Journal of the Korean Institute of Illuminating and Electrical Installation Engineers
• /
• v.13 no.2
• /
• pp.63-70
• /
• 1999

This paper is proposed a new method to deal with the optimized auto-tuning for the Pill controller which is used to the process-control in various fields. First of all, in this method, 1st order delay system with dead time which is modelled from the unit step response of the system is Pade-approximated, then initial values are determined by the Ziegler-Nichols method. So we can find the parameters of Pill controller so as to minimize the fuzzy criterion function which includes the maximum overshoot, damping ratio, rising time and settling time. Finally, after studying the parameters of Pill controller by Backpropagation of Neural-Network, when we give new K, L, T values to Neural-Network, the optimized parameter of Pill controller is found by Neural-Network Program.rogram.

• Computational Structural Engineering
• /
• v.8 no.1
• /
• pp.137-146
• /
• 1995

Structural stress under shock or impact load is varied with the lapse of time and the structural stress is called stress wave. Propagating longitudinal stress wave is studied in a 2-dimensional plate. A finite element program for elastic stress wave propagation is developed in order to investigate the shape of stress field at time increment. The longitudinal stress wave is generated by unit step function. According to the finite element analysis results, the longitudinal stress wave propagates to the similar direction of impact load and the front of stress wave propagates with the same speed as analytic solution and the shape of stress field is similar to that of analytic solution. The shear wave is occurred after the longitudinal stress wave and declined at an angle of 45 degrees compared with longitudinal stress wave and the speed of shear wave is about a half of the longitudinal stress wave. The intensity of shear wave is larger than that of longitudinal stress wave.

• Economic and Environmental Geology
• /
• v.28 no.5
• /
• pp.481-485
• /
• 1995

This paper presents a new approaching method to determine the velocity and geometry of shallow subsurface from seismic refraction events. After picking the first breaks from seismic refraction data, we assume that field refraction seismogram can be replaced by the unit delta function having time shift of first break. Time curves are generated by shooting ray tracing. The partial derivatives seismogram for a damped least squares method is computed analytically at each step of the forward ray tracing. The technique is successfully tested on synthetic and real data. It has the advantage of real full waveform inversion, which is robust at low frequency band even if the initial guess is far from the true model.

• Journal of The Korea Institute of Healthcare Architecture
• /
• v.23 no.3
• /
• pp.47-56
• /
• 2017

Purpose : This study is a basic research to establish guidelines for healthcare facility in Korean healthcare market. As a first step, it is a main purpose for making beginning point of ICU guideline to set a basic level of ICU facility size and quality by analysis current ICU facilities and existed domestic guidelines : law, criteria of healthcare insurance review and assessment service, credit of Korea institute for healthcare accreditation. Methods : First of all, the all requirements of existed guidelines are put together and summarized. The summary is compared with that of US, UK, and Australia to know its quality difference with international market. In addition, all hospitals in Korean market are classified based on total bed number and ICU bed number to know the most occupied facility size range in the market. Second of all, by comparing 15 ICUs' current setting of the public general hospital, a general condition of ICU facility would be extracted based on function, floor area per bed, services Results : 72.8% of hospitals in Korea are belonging below 500 beds hospital. Among them 200-299 beds hospitals occupied 35.3% and are shown as the most occupied hospital size. As 15 public general hospitals are analyzed, it is clear that the more bed size and services the more ICU area per bed. As a result it is sure that the 300~500 bed may be a clear line as a general ICU condition whose function and bed number relatively consistent in the range. Implications : to keep the qualified medical environment and contemporary hospital trend, the guideline as a minimum requirement keep naturally out from current healthcare settings and should reflect their limit to reconcile with the new trend in the market.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.20 no.3
• /
• pp.617-631
• /
• 2019

In the four-step demand model, a gravity mode is used most commonly at the trip distribution stage. The purpose of this study was to develop a new friction factor that can express the accessibility property as a single friction factor to compensate for the variable limits of the gravity model parameters (travel time, travel cost). To derive a new friction factor, a new friction factor was derived using the space syntax that can quantify the characteristics of the urban space structure, deriving the link-unit integration degree and then using the travel time and travel distance relationship. Calibration of the derived friction factor resulted in a similar level to that of the existing friction factor. As a result of verifying the various indicators, the explanatory power was found to be excellent in the short - and long - distance range. Therefore, it is possible to derive and apply the new friction factor using the integration index, which can complement the accessibility beyond the limit of the existing shortest distance, and it is believed to be more advantageous in future utilization.