• The Transactions of the Korean Institute of Electrical Engineers
• /
• v.39 no.8
• /
• pp.856-867
• /
• 1990

In this paper, the convergence and the learning accuracy of the back-propagation (BP) method in neural network are investigated by 1) analyzing the reason for decelerating the convergence of BP method and examining the rapid deceleration of the convergence when the learning is executed on the part of sigmoid activation function with the very small first derivative and 2) proposing the modified logistic activation function by defining, the convergence factor based on the analysis. Learning on the output patterns of binary as well as analog forms are tested by the proposed method. In binary output patter, the test results show that the convergence is accelerated and the learning accuracy is improved, and the weights and thresholds are converged so that the stability of neural network can be enhanced. In analog output patter, the results show that with extensive initial transient phenomena the learning error is decreased according to the convergence factor, subsequently the learning accuracy is enhanced.

• Journal of Nutrition and Health
• /
• v.31 no.4
• /
• pp.687-696
• /
• 1998

This study was performed to investigate effects of hesperidin and naringin on linpid peroxide formation and antioxidative enzyme activities in rats. Thiobarbituric acid reactive substance (TBARS) concentrations were measured in plasma and liver. Catalase, superoxide dismutase, and glutathione peroxidase activities were measured in erythrocyte and liver. Forty-nine male Sprague-Dauley rats weighing 275.3$\pm$3.3g were blocked into seven groups according to body weight and were raised fro four weeks on diets containing 0.25, 0.50 or 1.00%(w/w) hesperidin or naringin . Food intake, weight gain , food efficiency ratio, and weights of liver, kidney, spleen ,and epididymal fat pad were not significantly different among groups. In 0.50 and 1.00% naringin groups , plasma TBARS concentrations were significantly decreased with a dose response patter. In 0.25, 0.50 and 1.00% hesperidin groups, liver TBARS concentrations were significantly decreased without a dose dependent patter. Antiosidative enzyme activities in erythrocyte and liver were not significantly affected by type and amountof dietary bioflavonoid, but in the 1.00% hesperidin group, catalase, superoxide dismutase, and glutahione perosidase activities in linver showed a tendency to increase. In conclusion, naringin inhibited lipid peroxide formation with a dose response pattern in plasma without changing the activities of antioxidative enzymes. Hesperidin adminstration, regardless of the level in the diet, inhibited lipid peroxide formation in liver.

• 한국정보디스플레이학회:학술대회논문집
• /
• 2009.10a
• /
• pp.1450-1452
• /
• 2009

In this work, a plastic diffuser which has a surface similar to Sinusoidal wave profile for application of LC D backlighting is proposed. This new pattern diffuser is achieved the values of the transmission efficiency an d diffusion efficiency all over 70%, and it also has the value about gain over 1.1. Additionally, this new patter n diffuser has high uniformity when we put it in the LCD backlight module and its thickness is only 2mm. With these foregoing properties, this diffuser film can widely applied to LCD displays systems etc.

• Proceedings of the KSRS Conference
• /
• 2003.11a
• /
• pp.597-599
• /
• 2003

Moving pattern is as a kind of sequential pattern, which can be extracted from the large volume of location history data. This sort of knowledge is very useful in supporting intelligence to the LBS or GIS. In this paper, we proposed the continuous moving pattern mining approach in LBS platform and LBS Miner. The location updates of moving objects affect the set of the rules maintained. In our approach, we use the validity thresholds that indicate the next time to invoke the incremental pattern mining. The mining system will play a major role in supporting the various LBS solutions.

• Journal of Korea Technical Association of The Pulp and Paper Industry
• /
• v.41 no.1
• /
• pp.17-23
• /
• 2009

The influence of different forming elements on the local paper structure and the related paper properties was investigated in this study. Specifically, a conventional papermaking foil system and a velocity induced drainage (VID) system were compared. The study involved the analysis of the product samples obtained from the commercial machine trials. The paper samples produced with VID forming systems showed better formation. The deterministic patter in the local structural profile map of the Foil samples indicated the structure of foil samples was more supple after forming process and then easier to be marked by various fabrics such as wet pressing fabric. The higher bulk was observed in the VID samples, which resulted in higher scattering coefficient, lower ZDT strength, and higher bending stiffness.

• The Transactions of the Korean Institute of Power Electronics
• /
• v.11 no.1
• /
• pp.90-96
• /
• 2006

A novel high power factor drive of a single-phase switched reluctance motor (SRM) is researched. It achieves sinusoidal and near unity power factor input currents. The proposed SRM drive has one additional active switches. And a single-stage approach has a simple structure and low cost. A prototype to drive an SRM equipping a suitable encoder is designed to evaluate the proposed topology. The characteristics and validity of the proposed circuit is discussed with some simulations and experimental results.

• Journal of Korea Society of Digital Industry and Information Management
• /
• v.6 no.1
• /
• pp.1-12
• /
• 2010

Recently, large amount of information in IDS(Intrusion Detection System) can be un manageable and also be mixed with false prediction error. In this paper, we propose a data mining methodology for IDS, which contains uncertainty based on training process and post-processing analysis additionally. Our system is trained to classify the existing attack for misuse detection, to detect the new attack pattern for anomaly detection, and to define border patter between attack and normal pattern. In experimental results show that our approach improve the performance against existing attacks and new attacks,from 0.62 to 0.84 about 35%.

• 한국경영정보학회:학술대회논문집
• /
• 2008.06a
• /
• pp.153-170
• /
• 2008

Customers' needs change every moment. Profitability of stores can't be increased anymore with an existing standardized chain store management. Accordingly, a personalized store management tool needs through prediction of customers' preference. In this study, we propose a recommending procedure using dynamic customers' preference by analyzing the transaction database. We utilize self-organizing map algorithm and association rule mining which are applied to cluster the chain stores and explore purchase sequence of customers. We demonstrate that the proposed methodology makes an effect on recommendation of products in the market which is characterized by a fast fashion and a short product life cycle.

• Journal of Korean Medical classics
• /
• v.18 no.4 s.31
• /
• pp.145-154
• /
• 2005

The SangHanIlJeGeum(傷寒-提金) explain the six channel of disease. The Greater Yang controls the construction and defence, and governs the exterior of the body, which serves as the body's external barrier. Accordingly, many of the signs associated with greater yang appear in the early stages of disease. The essential features of greater yang disease are a pulse that is floating, headache, stiffness and pain of the head and nape, heateffusion, and aversion to cold. The main feature of Yang brightness disease is yang hyperactivity and heat exuberance. An evil can directly enter the yang brightness channel from the exterior, but it usually passes into the channel from the greater yang. Yang brightness disease is generally characterized by generalized heat effusion, spontaneous sweating, aversion not to cold but to heat, and a pulse that is large. Distinction is made between a heat patter and a repletion pattern.

• Journal of Electrical Engineering and information Science
• /
• v.3 no.1
• /
• pp.73-79
• /
• 1998

In this paper, a new traffic sign detection algorithm.. and a symbol recognition algorithm are proposed. For traffic sign detection, a dominant color transform is introduced, which serves as a tool of highlighting a dominant primary color, while discarding the other two primary colors. For symbol recognition, the curvilinear shape distribution on a circle centered on the centroid of symbol, called a circular pattern vector, is used as a spatial feature of symbol. The circular pattern vector is invariant to scaling, translation, and rotation. As simulation results, the effectiveness of traffic sign detection and recognition algorithms are confirmed, and it is shown that group of circular patter vectors based on concentric circles is more effective than circular pattern vector of a single circle for a given equivalent number of elements of vectors.