• Journal of the Korean Institute of Intelligent Systems
• /
• v.15 no.7
• /
• pp.846-851
• /
• 2005

In this paper, we propose Genetic Algorithms(GAs)-based Optimized Polynomial Neural Networks(PNN). The proposed algorithm is based on Group Method of Data Handling(GMDH) method and its structure is similar to feedforward Neural Networks. But the structure of PNN is not fixed like in conventional neural networks and can be generated in a dynamic manner. As each node of PNN structure, we use several types of high-order polynomial such as linear, quadratic and modified quadratic, and it is connected as various kinds of multi-variable inputs. The conventional PNN depends on the experience of a designer that select the number of input variables, input variable and polynomial type. Therefore it is very difficult to organize optimized network. The proposed algorithm leads to identify and select the number of input variables, input variable and polynomial type by using Genetic Algorithms(GAs). The aggregate performance index with weighting factor is proposed as well. The study is illustrated with tile NOx omission process data of gas turbine power plant for application to nonlinear process. In the sequel the proposed model shows not only superb predictability but also high accuracy in comparison to the existing intelligent models.

• Journal of the Korean Society of Food Science and Nutrition
• /
• v.32 no.3
• /
• pp.428-436
• /
• 2003

Epidemiological studies have observed a negative association between increased consumption of green and yellow vegetables and cancer incidence. These vegetables contain carotenoids, which are reported to exhibit anticarcinogenic effects. Overexpression of ErbB2 and ErbB3 genes is a frequent event in several human cancers. The present study was performed to determine whether $\alpha$-carotene, $\beta$-carotene, lutein, or lycopene inhibits cell growth and to assess such an effect is related to changes in the levels of the ErbB receptor family and tile ErbB3 receptor signaling pathway in HT-29 cells. HT-29 cells were cultured in serum-free medium in the presence of various concentrations (0~100 $\mu$M) of the individual carotenoids. $\alpha$ -Carotene and lycopene significantly inhibited cell growth in a dose-dependent manner, whereas lutein slightly inhibited cell growth and $\beta$-carotene increased cell growth. Lycopene is more potent than $\alpha$ -carotene in inhibiting HT-29 cell growth. Lycopene inhibited DNA synthesis and induced apoptosis of HT-29 cells. The ErbB3 ligand heregulin (HRG) increased cell growth but did not prevent the lycopene-induced inhibition of cell growth. Lycopene decreased ErbB2 protein levels in a dose-dependent manner. Immunoprecipitation/Western blot studies revealed that lycopene inhibited HRG-induced phosphorylation of ErbB3, recruitment of the 985 regulatory subunit of phosphatidylinositol 3-kinase (PI3K) to the ErbB3 receptor, and phosphorylation of Akt. These results indicate that downregulation of ErbB2/ErbB3/PI3K/Akt signaling may be one of the mechanisms by which lycopene inhibits HT-29 cell pro-liferation and induces apoptosis.

• Journal of the Korean Society of Food Science and Nutrition
• /
• v.34 no.9
• /
• pp.1407-1413
• /
• 2005

Korean households' expenditures on the eating -out have been steadily increasing except for the foreign exchange crisis period. This paper aimed to examine the eating-out expenditures of salary and wage earners's households by educational attainment of household head. We modeled the demand function of eating-out in terms of income and price, examining the responsiveness of eating-out demand to changes in income and price using econometric methods such as regression, rolling regression and impulse response. This paper showed that coefficients of income and price tended to decrease as the income rises in the long-run. Specifically, the higher the educational attainment of household head is, the smaller the coefficients of income and price we. The impulse response analysis also indicated that while price shocks decreased the demand for eating-out, income shocks tended to increase tile demand for eating-out in all the educational attainments. Furthermore income shocks were much greater and last longer than price shocks at all educational attainments of household heads.

• Journal of the Korean Society of Food Science and Nutrition
• /
• v.34 no.9
• /
• pp.1427-1433
• /
• 2005

This study investigated the quality characteristics of sponge cakes with addition of corn starch. The corn starch was added with 0 (control), 10, 20, 30, and $40\%$ substitution ratios by weight per wheat flour, followed by measuring physical and sensory characteristics of product during storage Periods. Batter specific gravity decreased with increasing substitution levels while viscosity increased. The volume tended to increase according to the addition of corn starch. As storage progresses, hardness and chewiness increased in all treatments. $30\%$ treatments showed lower values, compared to other treatments throughout the storage period. Adhesivness increased as the amount of corn starch increased. The results of sensory evaluation showed that the overall acceptability of sponge cakes containing $20\%\;and\;30\%$ corn starch were higher than that of control without corn starch. Therefore, it can be suggested that of corn starch $30\%$ added to tile sponge cake for the quality improvement and functional element.

• The Korean Journal of Physiology
• /
• v.3 no.1
• /
• pp.19-28
• /
• 1969

Skin temperatures on 10 sites and rectal temperature at every 10 minutes, oxygen consumption at every 20 minutes were measured on 18 male subjects (ages between 14 and 47 years) after exposure to cold air at $15^{\circ}C$ for two hours in a climatic room. Total body fat measured by means of a skinfold method and ratio of body surface area (S) to body volume (V), S/V, were utilized as basis of observations. Surface area was calculated after DuBois equation and body volume was calculated by our original formula. In influencing on the heat loss from the body core to the cold environment, % fat showed inverse relations, whereas, S/V ratio showed direct relations. Thus these two factors acted antagonistically on the body heat loss. Local skin temperatures showed negative correlations with skinfold thickness on the same site, nemaly, on chest, r=-.567; on back, r=-.507; and on upper arm, r=-.353. The other 7 skin sites showed low correlations with % fat. Minimum mean weighted skin temperature (MWST) showed a negative correlation (r=-.443) with % fat, and showed no correlation with S/V ratio. Oxygen consumption in the cold air at $15^{\circ}C$ increased from the first measurement at 20 minutes after exposure and maintained the same increasing trend up to 120 minutes. ${\Delta}T_R$ was greater in tile lean subjects who showed a greater % change in oxygen consumption. The antagonistic actions of % fat and S/V ratio on the heat loss were manifested by observations as follows: minimum rectal temperature was higher In fat subjects (r=.600) and lower in subjects with a greater S/V ratio (=-.582), ${\Delta}T_R$ was smaller in fat subjects (r=-.738) and greater in subjects with a greater S/V ratio (r=.618). Temperature difference between body core and skin surface (minimum rectal temperature minus minimum MWST) showed a positive correlation with % fat (r=.600) and a negative correlation with S/V ratio (r=-.881). Decrease in the mean body temperature and heat debt, respectively, showed negative correlations with % fat and positive correlations with S/V ratio.

• The Korean Journal of Physiology
• /
• v.4 no.1
• /
• pp.19-28
• /
• 1970

In order to investigate the changes in the temperature regulation mechanism following ingestion of red pepper, twenty healthy man were given 2 gms of powdered red pepper each in 100 ml of water at room temperature, $20^{\circ}C$, Simultaneous recordings of temperature and electrical resistance of the skin of forehead were taken by means of Physiograph, using appropriate transducers. Temperature of various spots such as the cheek, upper tack and the forearm was also measured by the telethermometer at interval of two minutes. The results obtained were as follows: 1. Seventeen out of 20 subjects showed immediate rise in the skin temperature of the forehead, and 11 of them manifested the oscillating pattern of the skin temperature. The average of the largest peak amplitude was $0.58{\pm}0.355^{\circ}C$. 2. Even those who failed to show the immediate rise in the temperature did not keep the quiescence over 9 minutes, and delayed responses were revealed. 3. The rise in tile skin temperature is the favourable sign for the heat discipating mechanism only because the sweat glands are activated with the concomittant rise in the temperature of the skin. 4. There was a preceding or coincide fall in the electrical resistance of the skin, and it was also attributed to the glandular activity. 5. At rather cool room temperature, $20^{\circ}C$, no visible sweat was seen during the period of observation. Nevertheless it was obvious that latent activation of the sweat glands was triggered and the. threshold was lowered. This situation imitates the acclimatized condition in the hot environment, and it is likely to increase the tolerance in tropical climate from the view point of temperature regulation.

• The Korean Journal of Quaternary Research
• /
• v.3 no.1
• /
• pp.87-101
• /
• 1989

The development of the Hantan river basin can be divided into three stages. The first stage include the ancient Hantan channel system prior to the Chongokni basalt which yield dates of about 0.6 mya from the K/Ar dating method. During this period the Baekuyri formation was formed. The Baekuyri formation is widely observed under the Chongokni basalt along the current river system. The second stage is the period when stream channels stayed on the top of the basalt plateau. Aggradation and deggradation were continued by meandering and braiding channel systems until major stream channel was formed. The currently remaining deposit on the top of the basalt was formed by lacustrine and fluvial systems in this period. During this period Pleistocene hominid was present on edge of water and flood plain and left Paleolithic material. This period was begun at the time of the final basalt flow dated about 300,000 BP. The third stage is designed for the time when the Hantan river channel was dropped down to a level from which the channel could not influence the top of the basalt any more No more deposit could be formed but erosion by surface water has been continued on the top of the basalt since then. The dropping of the Hantan river channel was probably not very long after the final flow of the basalt. Because of frost action and heavy concentrated precipitation in the basin area along with blocky and clumnar joint structure of the basalt, erosional process of the basalt is believed to have been carried out within a relatively short time. The lowering of the Hantan river channel was probably completed in a cycle of major fluctuation of world cimate. Also, the redclay on the top of the basalt is believed to have been formed during a warm period around 200,000 BP, which accords with the climatic change suggested above fair1y well. The Paleolithic materials in tile deposits fell accordingly into approximately same time period.

• Journal of Korean Society of Coastal and Ocean Engineers
• /
• v.19 no.2
• /
• pp.105-112
• /
• 2007

Sediment transport model based on the Lagrangian concept considering the grain size distribution(GSD) was setup and the change of the sediment diffusion range was analysed in the condition of considering and not considering the GSD. The GSD curve is assumed as the Log-normal distribution function in order to consider the GSD with respect to the Lagrangian concept and the random numbers, i.e. sediment particles, are generated based on the distribution function. The sediment particles is assumed as the spherical type and the random numbers based on the sediment weight is converted to the sediment diameters. Sediment transport patterns are analysed by the settling simulation, in which the settling velocity is computed by the van Rijn formulae and the horizontal diffusion coefficient is used as the constant parameter. The diffusion patterns are very similar to the patterns with GSD condition. The diffusion range defined as the range including 90%, 99% sediment weight of the total sediment weight, however, is larger than without considering GSD condition in 90%-option and shorter than with considering GSD condition in 99-option, respectively. The diffusion range is defined as tile p-percentage of the cumulative sediment weight region with reference to the 50% region, 90%- option, 99%-option, respectively.

• Korean Journal of Fisheries and Aquatic Sciences
• /
• v.2 no.2
• /
• pp.155-160
• /
• 1969

The soil textures and chemical compositions of Tapes Philippinarum beds were investigated in order to estimate the grade of the habitat soil of T. Philippinarum for the improvement of natural tidal flats. The coefficients of correlation between tile percentages of debris, coarse sand, fine sand, silt, and clay of the habitat soil, and the population density of T. Philippinarum were 0.78, 0.73, -0.42, -0.68, and -0.51 respectively. And the optimum levels of same were above $13.5\%,\;above\;23.0\%\;below\;41.5\%,\;below\;10.0\%$, and below $6.8\%$ respectively The $SiO_2\;Al_2O_3+Fe_2O_3$ ratio of the soil of T. Philippinarum beds showed the range of 5.09 to 5.60. The levels of soil pH, water holding capacity, organic matter, total nitrogen, available phosphorus, exchangeable magnesium were 6.15-6.60, $4.26-6.86\%$, $0.40-1.10\%$ 14.0-27.0 ppm, $0.041-0.394\%,\;and\;0.009-0.132\%$ respectively.

• Korean Journal of Fisheries and Aquatic Sciences
• /
• v.2 no.2
• /
• pp.161-172
• /
• 1969

Ten experimental tuna long line fishing operations were performed with the long lines A and B, in combination casting orders of BA and AB. The long line B differs from A in its additional main line length per basket and its greater depth of submersion (Table 1). The BA casting order results in an AB hauling order, so that the long line B is casted at an earlier time of a day and fishes for a longer period than A line (fig. 2). The number of fish caught was divided into subclasses in order to be subjected to $x^2-tests$ (Table 2). The first series of $x^2-tests$ were calculated using the hooking rates of two subclasses with other factors pooled (Table 3). The second series of $x^2-tests$ were calculated to obtain the interactions between two subclasses when other factors differed (Table 4). Attention was paid to the extremely small interactions as well as to the large interactions in order to find out whether a factor functions in the same way or in a contradictory way when the other factors involved differ. The test result is summarized in Table 5. The relationship between the hook casting time (relative to the sunrise time) and the hooking rates reveals the feeding behavior of particular species (fig. 3) A high hooking rate for yellowfin tuna shown in the period from one hour before sunrise to the moment of sunrise suggests that they feed actively in that period. Their poor hooking rate on earlier casted hooks suggests that they do not feed well under conditions of darkness and that they feed best while the hooks are sinking. Furthermore it is likely that yellowfin tuna maintain a shallow depth in the daytime, for the $x^2$ interaction between the line casting order BA and AB shows their persistant preferance of the long line A regardless of line casting order. Bigeye tuna show their feeding behavior under conditions of darkness, and show a preference for the sea's deeper layer. Alepisaurus borealis (Gill) show their feeding behavior under conditions of darkness more significantly than the bigeye tuna and their preferance for deep layers is also more significant. Marlins (mainly Makaria mazara) show a feeding behavior which is similar to the yellowfin tuna. The author emphasizes that certain variations of the hooking rates in relation to the line easting time were caused by an inadequate fishing depth as well as by the fish preference for a particular feeding time. When the part of the line casted in pre-dawn hours and hauled in the pre-sunset hours show a significantly lower hooking rate than other parts of the line, then it is concluded that tile hooks settled in a sea layer too deep for feeding.