• Journal of IKEEE
• /
• v.25 no.4
• /
• pp.664-671
• /
• 2021

This paper proposes a GPGPU structure that can reduce the number of operations and memory access by effectively applying a data reuse method to a convolutional neural network(CNN). Convolution is a two-dimensional operation using kernel and input data, and the operation is performed by sliding the kernel. In this case, a reuse method using an internal register is proposed instead of loading kernel from a cache memory until the convolution operation is completed. The serial operation method was applied to the convolution to increase the effect of data reuse by using the principle of GPGPU in which instructions are executed by the SIMT method. In this paper, for register-based data reuse, the kernel was fixed at 4×4 and GPGPU was designed considering the warp size and register bank to effectively support it. To verify the performance of the designed GPGPU on the CNN, we implemented it as an FPGA and then ran LeNet and measured the performance on AlexNet by comparison using TensorFlow. As a result of the measurement, 1-iteration learning speed based on AlexNet is 0.468sec and the inference speed is 0.135sec.

• Advances in pediatric surgery
• /
• v.12 no.2
• /
• pp.192-201
• /
• 2006

There are considerable controversies in the management of congenital diaphragmatic hernia. By 1997, early operation, routine chest tube on the ipsilateral side and maintainingrespiratory alkalosis by hyperventilation were our principles (period I). With a transition period from 1998 to 1999, delayed operation with sufficient resuscitation, without routine chest tube, and permissive hypercapnia were adopted as our practice. High frequency oscillatory ventilation (HFOV) and nitric oxide (NO) were applied, if necessary, since year 2000(period II). Sixty-seven cases of neonatal Bochdalek hernia from 1989 to 2005 were reviewed retrospectively. There were 33 and 34 cases in period I and II, respectively. The neonatal survival rates were 60.6 % and 73.5 %, respectively, but the difference was not significant. In period I, prematurity, low birth weight, prenatal diagnosis, inborn, and associated anomalies were considered as the significant poor prognostic factors, all of which were converted to nonsignificant in period II. In summary, improved survival was not observed in later period. The factors considered to be significant for poor prognosis were converted to be nonsignificant after change of the management principle. Therefore, we recommend delayed operation after sufficient period of stabilization and the avoidance of the routine insertion of chest tube. The validity of NO and HFOV needs further investigation.

• Journal of Practical Engineering Education
• /
• v.15 no.2
• /
• pp.381-388
• /
• 2023

Through artificial neural network education using spreadsheets, non-major undergraduate students can understand the operation principle of artificial neural networks and develop their own artificial neural network software. Here, training of the operation principle of artificial neural networks starts with the generation of training data and the assignment of correct answer labels. Then, the output value calculated from the firing and activation function of the artificial neuron, the parameters of the input layer, hidden layer, and output layer is learned. Finally, learning the process of calculating the error between the correct label of each initially defined training data and the output value calculated by the artificial neural network, and learning the process of calculating the parameters of the input layer, hidden layer, and output layer that minimize the total sum of squared errors. Training on the operation principles of artificial neural networks using a spreadsheet was conducted for undergraduate non-major students. And image training data and basic artificial neural network development results were collected. In this paper, we analyzed the results of collecting two types of training data and the corresponding artificial neural network SW with small 12-pixel images, and presented methods and execution results of using the collected training data for Orange machine learning model learning and analysis tools.

• Journal of Arbitration Studies
• /
• v.17 no.1
• /
• pp.77-95
• /
• 2007

This article focuses on the history and evolution of the US court's attitude towards unilateral arbitration and dispute resolution clauses, but also considers the practical approach of national courts to theses clauses. It goes on to consider some potential pitfalls in the operation of unilateral clauses, which should be borne in mind when developing a strategy for bringing or defending a claim which falls within the scope of a unilateral clause. There can be few objections to the general validity of unilateral arbitration clauses. The principle of party autonomy is the driving force behind international arbitration and, provided it is tolerably clear that the parties intended the arbitration clause to operate unilaterally, courts should be reluctant to interfere with the parties' agreement. There are also no persuasive public policy reasons why such clauses should not be upheld in commercial agreements. In addition to the issue of whether such unilateral clauses are permissible under certain law, it is important to be aware of how they should properly operate in practice, that is, useful guidance on the subject of the proper operation and effect of such clauses where they are intended to be used to enable a party to decide whether, and in what circumstances, a claim should be referred to court or to arbitration.

• Journal of the Korean Society for Aviation and Aeronautics
• /
• v.22 no.1
• /
• pp.22-31
• /
• 2014

Military UAV(Unmanned Aerial Vehicle) operated as a RC(Remotely Control) model level within the limit of the military special use airspace until now. However, the high and medium altitude of URA(Unmanned Reconnaissance Aircraft) which the ROKAF have been trying to import recently is at the UAV level and needs the criteria for the classified airspace flights. The required flight criteria includes operator location, mission operation limit, equipment, etc., which are the principle and standard applied based on the airspace use for UAV. Also, the general flight rules, visual flight rules, instrument flight rules are required in order to have to be applied to the actual flight. Besides, an appliance regulation needs to be arranged regarding two-way communication, ATC and communication issue, airspace and area in-flight between UAS( Unmanned Aircraft System) users. An operation of the UAV in the air significantly requires the guarantee of the aircraft's capacity, and also the standardized flight criteria. A safe and smooth use is ensured only if this criteria is applied and understood by the entire airspace users. For the purpose, a standardized military UAV flight operations criteria that is to be applied for each airspace by UAV is to be prepared through analysis of the present state, a legend UAV system, and a special character analysis.

• Journal of Chest Surgery
• /
• v.20 no.3
• /
• pp.536-543
• /
• 1987

The Fontan procedure was physiological correction which was initially applied to tricuspid atresia. We had used the modified Fontan operation in 8 cases at National Medical Center, Seoul, from Aug. 1984 to Oct. 1986. Age range was 20 months to 15 years [mean: 9 years] and male: female ratio was 5:3. 5 patients had tricuspid atresia [lb: 2 cases, Ic: 1 case, lib: 1 case, & llc: 1 case], 2 Patients had univentricular heart of left ventricular type, and one patient had transposition of great arteries with complete endocardial cushion defect. The operative principle was direct anastomosis between right atrium and pulmonary artery, whether main pulmonary artery or right pulmonary artery without any conduits. Postoperatively all patients needed high central venous pressure for adequate hemodynamic status in both survival [20-24 CmH2O] and mortality groups [20-24 CmH2O]. The fatal complications were as follows: empyema with bronchopleural fistula [1 case], bleeding tendency & brain damage [1 case], low cardiac output syndrome & acute renal failure [2 cases], and right to left shunt of unknown origin [1 case]. There were 5 hospital deaths; 3 of 5 tricuspid atresia patients, 1 of 2 univentricular heart patients, and 1 of 1 transposition of great arteries with complete endocar4ial cushion defect patient. The overall mortality was 62.5%. 3 survivors were nearly compatible with Choussat & Fontan criteria. Thus appropriate patient selection and experienced surgical technique were required for good results.

• Information and Communications Magazine
• /
• v.4 no.4
• /
• pp.505-524
• /
• 1987

Nowaday all the countries of the world have studied the various problems caused in operating their own ports efficiently. Ship delay in the port is attributal to the inefficient operation in the navigation aids, the cargo handling, the storage and transfer facilities, and to the inefficient allocation of gangs or to a bad service for ships. Among these elements the allocation of gangs is the predominating factor in minimizing ship's turn round time. At present, in the case of Pusan Port. the labour union and stevedoring companies allocate gangs in every hatches of ships by a rule of thumb, just placing emphasis on minimizing ship's turn round time, without applying the principle of allocation during the cargo handling. Owing to this the efficiency of the cargo handling could not be expected to be maximized and this unsystematic operation result in supplying human resources of much unnecessary surplus gangs. Therefore in this paper the optimal size and allocation of gangs for minimizing the ship's turn round time is studied and formularized. For the determination of the priority for allocation the evaluation function, namely $F=PHi^{n}{\times}(W+H)$, can be obtained; where, PHI : Principal Hatch Index W : Total Cargo Weight represented in Gang-Shifts H : Total Number of Ship's hatches and also for the optimal size of gangs the average number of gang allocated per shift (Ng), namely Ng=W/PHI, is used. The proposed algorithm is applied to Pusan Port and its validity is verified.

• Journal of IKEEE
• /
• v.22 no.2
• /
• pp.499-505
• /
• 2018

Since feedback field-effect transistors (FBFETs) have ideal switching characteristics resulting from feedback phenomenon caused by electrons and holes in the channel region, the researches about FBFET devices have been proposed and demonstrated worldwide recently. The device operated with novel principle can operate as a switching electronic device. Besides, because the hysteresis characteristics of the device by accumulated electrons and holes in channel region can be also utilized for memory applications, its application range is wide. In this paper, we cover various device structures of FBFET proposed until now and their operation mechanism, and then look into their applicable fields.

• Proceedings of the KIEE Conference
• /
• 2005.11a
• /
• pp.223-225
• /
• 2005

An external electrode fluorescent lamps (EEFLs) have the advantage of a long lifetime in the early stages of the study on plasma discharge, interest in the lamp continues. Studies on the operation of external electrode fluorescent lamps have focused mainly on its use of a type of high frequency (MHz). By performing high brightness using a square wave operation method with the low frequency below 100kHz, which is applied to a narrowed tube type lamp that has several mm of lamp diameter, an EEFL presented the possibility of using it as a light source for backlights. However, because an EEFL generates plasma using wall charges, which considers the impedance characteristics of glass based on the structural principle in discharge, it can be significantly affected by frequency. Thus, this study verifies the change in the characteristics of electromagnetic fields according to the change in frequency through a Maxwell's electromagnetic field simulation and examines the relationship between the change in the EEFL frequency and brightness by measuring the optical characteristics. In addition, the characteristics of the transformation of energy orbits were verified by investigating the characteristics of the wavelength according to the change in frequency through the OES.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.56 no.2
• /
• pp.355-360
• /
• 2007

The recent TFT-LCD Trend that is done large size gradually. As size of monitor great, though problem happens, it is that consumer's request which it makes monitor combined TV function. Monitor and TV are no difference externally greatly, but define difference happens as for backlight. An external electrode fluorescent lamp (EEFL) has an advantage of a long lifetime in the early stages of the study on plasma discharge, interest in the lamp continues. Researches on the operation of external electrode fluorescent lamps have focused mainly on its use of a type of high frequency (MHz). By performing high Luminance using a square wave operation method with the low frequency below 100kHz, which is applied to a narrowed tube type lamp that has several mm of lamp diameter, EEFL presented the possibility of using it as a light source for back-lights. However, because EEFL generates plasma using wall charges, which considers the impedance characteristics of glass based on the structural principle in discharge, it can be significantly affected by frequency. Thus, this study verified the change in the characteristics of electromagnetic fields according to the change in frequency through a Maxwell electromagnetic field simulation and examined the relationship between the change in the EEFL frequency and Luminance by measuring the optical characteristics.