• 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.

• Korean Journal of Remote Sensing
• /
• v.36 no.5_2
• /
• pp.929-937
• /
• 2020

In this manuscript, the UNet++ model, which is one of the representative deep learning techniques for semantic segmentation, was used to detect changes in temporal satellite images. To analyze the learning results according to various loss functions, we evaluated the change detection results using trained UNet++ models by binary cross entropy and the Jaccard coefficient. In addition, the learning results of the deep learning model were analyzed compared to existing pixel-based change detection algorithms by using WorldView-3 images. In the experiment, it was confirmed that the performance of the deep learning model could be determined depending on the characteristics of the loss function, but it showed better results compared to the existing techniques.

• Journal of Korean Society for Quality Management
• /
• v.38 no.4
• /
• pp.587-592
• /
• 2010

This paper deals with the performance evaluation of determining production size specifications. A customer who does not find the size specification he or she wants may purchase rather a larger or smaller one, but the purchasing desire decreases as the difference between the required and the prepared sizes increases. Introducing a generalized quality loss function which reflects how much the purchasing desire changes according to the difference, Park and Kim(1992) formulated a mathematical model for determining the size specifications so as to minimize the expected loss. Afterward the model has been applied to the determination of sizing system for mail order clothing and brassiere (Lee and Choi, 1996; Chun, et.al., 1996). The performance of the size determination method proposed by Park and Kim is evaluated in this paper. Usually the intervals between two successive size specifications are determined to be equal, but the size determination method compares favorably with the equidistance case, and more favorably if the population distribution is more skewed.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.9
• /
• pp.2882-2903
• /
• 2022

Image super-resolution (SR) processing is of great value in the fields of digital image processing, intelligent security, film and television production and so on. This paper proposed a densely connected deep learning network based on cascade architecture, which can be used to solve the problem of super-resolution in the field of image quality enhancement. We proposed a more efficient residual scaling dense block (RSDB) and the multi-channel cascade architecture to realize more efficient feature reuse. Also we proposed a hybrid loss function based on L₁ error and L_∞ error to achieve better L_∞ error performance. The experimental results show that the overall performance of the network is effectively improved on cascade architecture and residual scaling. Compared with the residual dense net (RDN), the PSNR / SSIM of the new method is improved by 2.24% / 1.44% respectively, and the L_∞ performance is improved by 3.64%. It shows that the cascade connection and residual scaling method can effectively realize feature reuse, improving the residual convergence speed and learning efficiency of our network. The L_∞ performance is improved by 11.09% with only a minimal loses of 1.14% / 0.60% on PSNR / SSIM performance after adopting the new loss function. That is to say, the L_∞ performance can be improved greatly on the new loss function with a minor loss of PSNR / SSIM performance, which is of great value in L_∞ error sensitive tasks.

• Proceedings of the KSME Conference
• /
• 2007.05b
• /
• pp.2414-2419
• /
• 2007

A rectangular fin with variable fin height, fin length and surrounding temperature is analyzed using a one-dimensional analytical method. Both the heat loss from a rectangular fin with non-insulated fin tip and that with insulated fin tip are presented as a function of the fin height, fin tip length and the convection characteristic number. The relative error in the heat loss of these two cases is also given as a function of the same variables. One of the results shows that the trend of heat loss for both cases with the variation of given variables is similar even though the relative error increases as the shape of the fin becomes shorter and fatter.

• Journal of Korean Society for Quality Management
• /
• v.31 no.4
• /
• pp.164-175
• /
• 2003

This paper investigates bootstrap confidence intervals of the process capability index(PCI) based on the expected loss derived from the empirical distribution function(EDF). The PCI based on the expected loss is too complex to derive its confidence interval analytically, so the bootstrap method is a good alternative. We propose three types of the bootstrap confidence interval; the standard bootstrap(SB), the percentile bootstrap(PB), and the acceleration biased­corrected percentile bootstrap(ABC). We also perform a comprehensive simulation study under various process distributions, in order to compare the accuracy of the coverage probability of the bootstrap confidence intervals. In most cases, the coverage probabilities of the bootstrap confidence intervals from the EDF PCI turned out to be more accurate than those from the PCI based on the normal distribution. It is expected that the bootstrap confidence intervals from the EDF PCI can be utilized in real processes where the true distribution family may not be known.

• New & Renewable Energy
• /
• v.3 no.1 s.9
• /
• pp.13-19
• /
• 2007

A triangular fin with variable fin base thickness and base height is analyzed and optimized for the fixed fin base height using a two-dimensional analytical method. At the middle of the fin length, the variation of the temperature along the fin height is listed. The influences of the fin length, base thickness and base height on the heat loss and fin efficiency are analyzed, The optimum heat loss, corresponding optimum efficiency and optimum fin length as a function of the fin base thickness are presented. The optimum heat loss and optimum fin tip length as a function of the convection characteristic number are represented.

• Journal of Korean Society of Water and Wastewater
• /
• v.23 no.1
• /
• pp.107-112
• /
• 2009

This study was conducted for verification and systematization of estimation method about the headloss using CFD(Computational Fluid Dynamics). Head loss which happens between the inlet and outlet of in-line mixer can be a major factor for the design and construction. Also, this Case studies about the sensitivity related to the velocity in the piping system. As result, program's default calculation function was used to get each side's total pressure and the differential of each total pressure could be defined as head loss from in-line mixer. In the case of adopting pipe surface friction factor and geometry loss, Calculation residual can be much more reduced. It was found that residual of value between CFD method and field test ranged about 3 through 18 precent.

• Korean Journal of Air-Conditioning and Refrigeration Engineering
• /
• v.16 no.8
• /
• pp.683-690
• /
• 2004

A modified rectangular fin is analyzed by two-dimensional analytic method and finite difference method. Relative error of heat loss from the modified rectangular fin between analytic method and finite difference method is presented. Comparisons of fin effectiveness and heat loss between a modified rectangular fin and a plane rectangular fin are made as a function of the non-dimensional fin length and wing height for different positions of wings by using analytic method. The ratio of the incremental rate of heat loss to that of the area of a modified rectangular fin is shown as a function of the wing height. One of the results shows that performance of a modified fin is more improved as the wing approaches the fin root.

• Proceedings of the KIEE Conference
• /
• 1993.07b
• /
• pp.968-970
• /
• 1993

In this paper, the loss components of IPMSM(Inteiror Permanent Magnet Synchronous Motor) is derived. To maximize the efficiency of the motor, a design method that optimizes the design variables is proposed Objective function consists of stator winding loss, core loss, and mechanical loss. Simulated annealing is used as the optimization method which is appropriate for finding the global minimum of nonlinear function with many local minima. Through the simulation of the motor characteristics, the prominence of the proposed design method is verified.