• Proceedings of the Korea Water Resources Association Conference
• /
• 2007.05a
• /
• pp.855-859
• /
• 2007

It is essential to obtain accurate and highly reliable streamflow data for water resources planning, evaluation and management as well as design of hydraulic structures. A new discharge estimation method, which is named 'non-dimensional velocity distribution and index-velocity method,' was proposed in this research. This method showed very close channel discharges which were calculated with the exiting velocity-area method. When velocity-area method is used to estimate channel discharge, it is required to observe point velocities at every desired point and vertical using a current meter like Price-AA. However 'non-dimensional velocity distribution and index-velocity method' is used, it become optional to observe point velocities at every desired point and vertical. But this method can not be applied for the cases of very complex and strongly asymmetric channel cross-sections because non-dimensional velocity distribution by entropy concept may be quite biased from that of natural rivers.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.8 no.6
• /
• pp.1560-1565
• /
• 2007

In this study, Chiu's velocity distribution equation recently developed from the probability and entropy concepts is used to establish a linkage between the mean velocity obtained from the Manning's equation and the corresponding velocity distribution in a channel cross section. The linkage to be established enables computing the velocity distribution along with the mean velocity, from simple hydraulic data such as Manning's n, hydraulic radius and channel slope irrespective of including sediment or not.

• Communications for Statistical Applications and Methods
• /
• v.22 no.5
• /
• pp.487-494
• /
• 2015

A calculation of the Kullback-Leibler information of consecutive order statistics is complicated because it depends on a multi-dimensional integral. Park (2014) discussed a representation of the Kullback-Leibler information of the first r order statistics in terms of the hazard function and simplified the r-fold integral to a single integral. In this paper, we first express the Kullback-Leibler information in terms of the reversed hazard function. Then we establish a generalized result of Park (2014) to an arbitrary consecutive order statistics. We derive a single integral form of the Kullback-Leibler information of an arbitrary block of order statistics; in addition, its relation to the Fisher information of order statistics is discussed with numerical examples provided.

• Journal of Soil and Groundwater Environment
• /
• v.11 no.3
• /
• pp.52-58
• /
• 2006

Kim and Lee (2005) suggested Minimum Entropy Deconvolution (MED) to estimate the temporal sequence of the relative recharge. However this study by Kim and Lee (2005) was just related to the verification of the conceptual approach with MED. In this study, we try to characterize the applicability of MED in the case of spatially heterogeneous recharge (distance from recharge area). Simulated results were recorded with some specific sampling points. Estimated results from this study show higher than 0.8 in cross-correlation with the original recharge sequence. In addition, the physical and mathematical meanings of the applied filter length was also investigated. It was revealed that the length of filter is highly related to the spatial distance between recharge area and the monitoring site, and the apparent shape of hydraulic head change.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.21 no.2
• /
• pp.189-194
• /
• 2020

This paper analyzes the gradient descent method, which is the one most used for learning neural networks. Learning means updating a parameter so the loss function is at its minimum. The loss function quantifies the difference between actual and predicted values. The gradient descent method uses the slope of the loss function to update the parameter to minimize error, and is currently used in libraries that provide the best deep learning algorithms. However, these algorithms are provided in the form of a black box, making it difficult to identify the advantages and disadvantages of various gradient descent methods. This paper analyzes the characteristics of the stochastic gradient descent method, the momentum method, the AdaGrad method, and the Adadelta method, which are currently used gradient descent methods. The experimental data used a modified National Institute of Standards and Technology (MNIST) data set that is widely used to verify neural networks. The hidden layer consists of two layers: the first with 500 neurons, and the second with 300. The activation function of the output layer is the softmax function, and the rectified linear unit function is used for the remaining input and hidden layers. The loss function uses cross-entropy error.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.25 no.8
• /
• pp.1046-1052
• /
• 2021

Artificial intelligence is being applied in various industrial fields to the development of the fourth industry and the construction of high-performance computing environments. In the medical field, deep learning learning such as cancer, COVID-19, and bone age measurement was performed using medical images such as X-Ray, MRI, and PET and clinical data. In addition, ICT medical fusion technology is being researched by applying smart medical devices, IoT devices and deep learning algorithms. Among these techniques, medical image-based deep learning learning requires accurate finding of medical image biomarkers, minimal loss rate and high accuracy. Therefore, in this paper, we would like to compare and analyze the performance of the Cross-Entropy function used in the image classification algorithm of the loss function that derives the loss rate in the chest X-Ray image-based deep learning learning process.

• Transactions of Materials Processing
• /
• v.29 no.3
• /
• pp.163-171
• /
• 2020

We present a multi-step cold drawing for a non-equiatomic Co₁₀Cr₁₅Fe₂₅Mn₁₀Ni₃₀V₁₀ high entropy alloy (HEA) with a simple face-centered cubic (FCC) crystal structure. The distribution of strain in the cold-drawn Co₁₀Cr₁₅Fe₂₅Mn₁₀Ni₃₀V₁₀ HEA wires was analyzed by the finite element method (FEM). The effective strain was expected to be higher as it was closer to the surface of the wire. However, the reverse shear strain acted to cause a transition in the shear strain behavior. The critical effective strain at which the shear strain transition behavior is completely shifted was predicted to be 4.75. Severely cold-drawn Co₁₀Cr₁₅Fe₂₅Mn₁₀Ni₃₀V₁₀ HEA wires up to 96% of the maximum cross-sectional reduction ratio were successfully manufactured without breakage. With the assistance of electron back-scattering diffraction and transmission electron microscope analyses, the abundant deformation twins were found in the region of high effective strain, which is a major strengthening mechanism for the cold-drawn Co₁₀Cr₁₅Fe₂₅Mn₁₀Ni₃₀V₁₀ HEA wire.

• Proceedings of the Korean Society of Soil and Groundwater Environment Conference
• /
• 2003.09a
• /
• pp.574-578
• /
• 2003

There are so many methods to estimate the groundwater recharge. These methods can be categorized into four groups. First groupis related to the water balance analysis, second group is concerned with baseflow/springflow recession, and third group is interested in some types of tracers; environmental tracers and/or temperature profile. The limitation of these types of methods is that the estimated results of recharge are presented in the form of an average over some time period. Forth group has a little different approach. They use the time series data of hydraulic head and specific yield evaluated from field test, and the results of estimation are described in the sequential form. But their approach has a serious problem. The estimated results in forth typeof methods are generally underestimated because they cannot consider the discharge phase of water table fluctuation coupled with the recharge phase. Ketchum el. at. (2000) proposed calibrated method, considering recharge- and discharge-coupled water table fluctuation. But the dischargeis considered just as the areal average with discharge rate. On the other hand, there are many methods to estimate the source wavelet with observed data set in geophysics/signal processing and geophysical methods are rarely applied to the estimation of groundwater recharge. The purpose this study is the evaluation of the applicability of one of the geophysical method in the estimation of sequential recharge rate. The applied geophysical method is called minimum entropy deconvolution (MED). For this purpose, numerical modeling with linearized Boussinesq equation was applied. Using the synthesized hydraulic head through the numerical modeling, the relative sequenceof recharge is calculated inversely. Estimated results are very concordant with the applied recharge sequence. Cross-correlations between applied recharge sequence and the estimated results are above 0.985 in all study cases. Through the numerical test, the availability of MED in the estimation of the recharge sequence to groundwater was investigated

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.10 no.5
• /
• pp.57-62
• /
• 2010

Recently, decision tree techniques have been studied in terms of quick searching and extracting of massive data in medical fields. Although many different techniques have been developed such as CART, C4.5 and CHAID which are belong to a pie in Clermont decision tree classification algorithm, those methods can jeopardize remained data by the binary method during procedures. In brief, C4.5 method composes a decision tree by entropy levels. In contrast, CART method does by entropy matrix in categorical or continuous data. Therefore, we compared C4.5 and CART methods which were belong to a same pie using breast cancer data to evaluate their performance respectively. To convince data accuracy, we performed cross-validation of results in this paper.

• Advances in nano research
• /
• v.12 no.2
• /
• pp.223-239
• /
• 2022

Object detection has always been to pursue objects with particular properties or representations and to predict details on objects including the positions, sizes and angle of rotation in the current picture. This was a very important subject of computer vision science. While vision-based object tracking strategies for the analysis of competitive videos have been developed, it is still difficult to accurately identify and position a speedy small ball. In this study, deep learning (DP) network was developed to face these obstacles in the study of tennis motion tracking from a complex perspective to understand the performance of athletes. This research has used CNN-VGG 16 to tracking the tennis ball from broadcasting videos while their images are distorted, thin and often invisible not only to identify the image of the ball from a single frame, but also to learn patterns from consecutive frames, then VGG 16 takes images with 640 to 360 sizes to locate the ball and obtain high accuracy in public videos. VGG 16 tests 99.6%, 96.63%, and 99.5%, respectively, of accuracy. In order to avoid overfitting, 9 additional videos and a subset of the previous dataset are partly labelled for the 10-fold cross-validation. The results show that CNN-VGG 16 outperforms the standard approach by a wide margin and provides excellent ball tracking performance.