• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.39 no.4
• /
• pp.462-469
• /
• 2002

While the conventional K-means algorithms use a fixed weight to design a vector codebook for all learning iterations, the proposed method employs a variable weight for learning iterations. The weight value of two or more beyond a convergent region is applied to obtain new codevectors at the initial learning iteration. The number of learning iteration applying a variable weight must be decreased for higher weight value at the initial learning iteration to design a better codebook. To enhance the splitting method that is used to generate an initial codebook, we propose a new method, which reduces the error between a representative vector and the member of training vectors. The method is that the representative vector with maximum squared error is rejected, but the vector with minimum error is splitting, and then we can obtain the better initial codevectors.

• Transactions of the Korean Society of Automotive Engineers
• /
• v.5 no.3
• /
• pp.182-190
• /
• 1997

Recently to develope an automobile with better properties, much researches and investments are executed in many countries. In this paper, it is intended to minimize the weight of an engine block without changing the dynamic characteristics. The weight minimization is executed by the sensitivity of the natural frequency of the engine block. To decrease the engine weight, much less thickness than the original thickness of the engine is selected to initial value and the structure modification is performed to recover the dynamic characteristics of the engine. Here, the original thickness of the engine is 8mm and the initial thickness is selected to 5mm, 6mm and the number if the natural frequencies fitted are 2, 6, 7, respectively. As the results, it is found that; (1) the weight of each case could be reduced without changing the objective natural frequencies. Specially, in the case of fitting 2 natural frequencies with 5mm initial thickness the weight could be reduced to 4.21kg(23.3% for engine weight). (2) according to the driving frequency range of engine, the weight minimization could be performed choicely, (3) improving a vibration characteristics of a orignal structure, the weight minimization could be performed.

• Journal of Korean Medicine for Obesity Research
• /
• v.1 no.1
• /
• pp.21-32
• /
• 2001

Objectives: The aim of this study was to the effect of very low calorie diet with Chegameuiintang to treat obese patients . Methods: We have evaluated 54 obese patients during very low calorie diet with Chegameuiintang weight-reducing regime by determining the body composition before and 1 month, 2 months after treatment by Bioelectrical impedence analysis . Results: In the change of body composition by the treatment, mean weight reduction was 11.59kg(15.57% of initial weight), mean weight reduction after 1month was 7.68kg( 10.31% of initial weight), mean weight reduction after 2 months was 3.91 kg(5.86% of weight after 1 month). Mean reduction in total fat mass was 8.03kg(2914% of the initial fat mass) which constitutes 69.28%. Mean reduction in lean body mass was 3.56kg(7.59%) Conclusions: This study has demonstrated that the compound therapy of very low calorie diet and Chegameuiintang (Tiganyiyiren-tang) is the good way of weight reduction, which reduces total fat mass and preserves lean body mass.

• The Journal of Korean Medicine
• /
• v.37 no.3
• /
• pp.62-73
• /
• 2016

Objectives: Dropout and weight regain are common problems in most obesity treatments. The purpose of this study was to review previously published study results of the predictive factors associated with dropout during weight loss treatment and weight loss maintenance after successful weight loss. Methods: Authors searched for the articles related to dropout and weight loss maintenance, published from 2007 to 2016 found on Pubmed, Scopus, RISS, and KISS. A total of 19 articles were finally selected. From the study results, unchangeable and changeable predictors were extracted, and these predictors were examined according to dropout and weight loss maintenance categories. Results: The unchangeable predictors of dropout were younger age, lower education level and female, whereas the changeable predictors of dropout were lower initial weight loss, symptoms of depression and body dissatisfaction. The strongest factor for predicting the dropout was initial weight loss. The unchangeable predictors of weight loss maintenance were old age, male and family history of obesity, whereas the changeable predictors of weight loss maintenance were regular exercise, dietary restraint, self-weighing and low depressive symptoms. Initial weight loss, depressive symptoms, body image, dietary restraint, physical activity, weight loss expectation and social support were considered to be dominant factors for weight loss treatments. Conclusions: Our review results suggest that unchangeable and changeable predictors of dropout and weight loss maintenance should be carefully examined during treatments of obesity.

• 한국생물공학회:학술대회논문집
• /
• 2001.11a
• /
• pp.731-734
• /
• 2001

An efficient fermentation strategy for the high level production of ultra-high-molecular weight poly(3-hdyroxybutyrate) (PHB) was developed. Although the cell and PHA concentrations obtained by flask cultures at different initial pH (6.0 or 6.9) were almost same level, the molecular mass of PHB produced were quite different along with the initial pH. When a recombinant Escherichia coli XL1-Blue harboring pJC2 containing the Alcaligenes latus PHB biosynthesis genes was cultivated in flask culture (pH 6.0), the PHB having a very high molecular weight of 22 MDa could be produced while only below 1 MDa at initial pH 6.9. The ultra-high-molecular weight PHB could be synthesized to high concentration of 89.8 g/L resulting in the PHB productivity of 2.07 g/L-h by simple fed-batch culture. In this study, we report that PHB having various molecular mass can be produced by employing metabolically engineered E. coli strains harboring the plasmids of different copy numbers containing the A. latus phbCAB genes.

• Journal of the Korean Ceramic Society
• /
• v.29 no.9
• /
• pp.711-718
• /
• 1992

Pb(Fe1/2Nb1/2)O3 was prepared by the molten salt synthesis method using an equimolar mixture of NaCl-KCl. Initial particle size could be controlled by varying the weight ratio of the NaCl-KCl to raw materials from 0.1 to 1.0, and the initial particle size effects on the sintering and dielectric properties of Pb(Fe1/2Nb1/2)O3 were investigated at the sintering temperature range from 90$0^{\circ}C$ to 105$0^{\circ}C$. As the weight ratio of salt increased, the average particle size decreased and the particle size distributions tended to narrow. As the initial particle size decreased, the linear shrinkage and density increased due to the promotion of densification. Dielectric constant increased with decreasing the initial particle size resulting from the increase of density and grain size.

• Journal of Microbiology and Biotechnology
• /
• v.12 no.1
• /
• pp.1-7
• /
• 2002

The effects of DO and pH on the mass production of pullulan with high molecular weight and the morphology of A. pullulans ATCC 42023 were evaluated. A. pullulans showed a maximum production of pullulan (11.98 g/l) when the initial pH of the culture broth was 6.5 in a shake-flask culture. In a batch culture, the mixture of a yeast-like and mycelial cell forms was found at a pH of 4.5, and the maximum production of pullulan (13.31 g/l) was obtained. However, a high proportion of high molecular weight pullulan (M.W.>2,000,000) was produced at a pH of 6.5, with a yeast-like morphology. The maximum pullulan production yield ($51\%$) was obtained at a pH noncontrol (initial pH 6.5) and DO control (above $50\%$) condition. Pullulan degrading enzyme was activated when the pH of the broth was lower than 5.0 and the portion of low molecular weight pullulan was increased. The formation of a black pigment was observed at an initial stationary phase, at 40 h of fermentation. Therefore, the fermentation should be carried out in a pH noncontrol (initial pH of 6.5) and DO control (above $50\%$) condition, and should be harvested before reaching the stationary phase (around 40 h) for the production of high molecular weight pullulan.

• Journal of the Korean Society for Precision Engineering
• /
• v.31 no.5
• /
• pp.411-416
• /
• 2014

In this research we present design optimization methods for a vehicle-mounted satellite antenna positioner. Our initial antenna positioner was conservatively designed to satisfy a worst case scenario where wind blew across the positioner at the speed of 120 km/h. Investigating stresses and safety based on Finite Element Methods (FEM), we find reflector support frames can be optimized to significantly reduce the weight of the positioner system. Thus, we optimize the reflector support frame from the given initial design while considering weight, maximum stress, maximum allowable deflection, cross section, and thickness. As a result, Shape C and the thickness of 2 mm are determined for the cross section of the reflector support frame. Applying this result, the weight of the new antenna positioner is 57.343 kg, which is decreased by 10.74% compared to the initial conservative design.

• Proceedings of the KSME Conference
• /
• 2004.11a
• /
• pp.817-822
• /
• 2004

This paper provides a global path planning method using self-organizing feature map which is a method among a number of neural network. The self-organizing feature map uses a randomized small valued initial weight vectors, selects the neuron whose weight vector best matches input as the winning neuron, and trains the weight vectors such that neurons within the activity bubble are move toward the input vector. On the other hand, the modified method in this research uses a predetermined initial weight vectors, gives the systematic input vector whose position best matches obstacles, and trains the weight vectors such that neurons within the activity bubble are move toward the input vector. According to simulation results one can conclude that the modified neural network is useful tool for the global path planning problem of a mobile robot.

• Journal of the Korean Society for Precision Engineering
• /
• v.23 no.3 s.180
• /
• pp.94-101
• /
• 2006

A global path planning method using self-organizing feature map which is a method among a number of neural network is presented. The self-organizing feature map uses a randomized small valued initial weight vectors, selects the neuron whose weight vector best matches input as the winning neuron, and trains the weight vectors such that neurons within the activity bubble are moved toward the input vector On the other hand, the modified method in this research uses a predetermined initial weight vectors of 1-dimensional string and 2-dimensional mesh, gives the systematic input vector whose position best matches obstacles, and trains the weight vectors such that neurons within the activity bubble are moved toward the input vector. According to simulation results one can conclude that the modified neural network is useful tool for the global path planning problem of a mobile robot.