• Publications of The Korean Astronomical Society
• /
• v.32 no.3
• /
• pp.421-438
• /
• 2017

Through the short-period variability survey program, we obtained time-series BV CCD images for $1.5^{\circ}{\times}1.0^{\circ}$ region around the young open cluster NGC 457. As a result, we have detected 61 variable stars including 31 new ones after checking light curves of all stars by eyes. The 61 variable stars were included 14 ${\delta}$ Scuti variable stars, a ${\beta}$ Cephei variable star, 10 variable Be and slowly pulsating B stars, 13 eclipsing binary stars, 21 semi-long periodic or slow irregular variables and an RR Lyrae variable star, respectively. Many variable B-type stars were known through a well-defined zero-age main sequence to the ${\beta}$ Cepheid region of NGC 457. Most of the variable B-type stars found this paper were known variable stars. But, 11 out of 14 ${\delta}$ Scuti variable stars were newly discovered. The new variable stars except for ${\delta}$ Scuti stars were 4 variable B-type stars, 5 eclipsing binaries and 11 semi-long periodic or slow irregular variables. We have performed frequency analysis for all ${\delta}$ Scuti stars, a ${\beta}$ Cepheid star and an RR Lyrae star.

• Journal of Welding and Joining
• /
• v.23 no.6
• /
• pp.43-48
• /
• 2005

This work was carried out to obtain sound welds and to select a most suitable binary metal joint among three different dissimilar binary metal combinations such as Zr-4/Ta, Mo/Ta and Ti/Ta(seal tube/sensor sheath) joints fur the instrumented nuclear fuel irradiation test. To do this, Taguchi experimental method was employed to optimize the experimental data. In addition, metallography, micro-focus x-ray radiography and hardness test were conducted to examine the welds. From the weld bead appearance, penetration depth and bead width as well as weld defects standpoint, Zr-4/Ta joint is suggested for the circumferential joining between a seal tube and a sensor sheath. The optimized welding parameters based on Zr-4/Ta joint are suggested as well.

• Journal of Astronomy and Space Sciences
• /
• v.33 no.3
• /
• pp.185-196
• /
• 2016

New multiband BVRI light curves of NSVS 1461538 were obtained as a byproduct during the photometric observations of our program star PV Cas for three years from 2011 to 2013. The light curves indicate characteristics of a typical W-subtype W UMa eclipsing system, displaying a flat bottom at primary eclipse and the O'Connell effect, rather than those of an Algol/b Lyrae eclipsing variable classified by the northern sky variability survey (NSVS). A total of 35 times of minimum lights were determined from our observations (20 timings) and the SuperWASP measurements (15 ones). A period study with all the timings shows that the orbital period may vary in a sinusoidal manner with a period of about 5.6 yr and a small semi-amplitude of about 0.008 day. The cyclical period variation can be interpreted as a light-time effect due to a tertiary body with a minimum mass of 0.71 M_⊙. Simultaneous analysis of the multiband light curves using the 2003 version of the Wilson-Devinney binary model shows that NSVS 1461538 is a genuine W-subtype W UMa contact binary with the hotter primary component being less massive and the system shows a low mass ratio of q(m_c/m_h)=3.51, a high orbital inclination of 88.7°, a moderate fill-out factor of 30 %, and a temperature difference of ΔT=412 K. The O'Connell effect can be similarly explained by cool spots on either the hotter primary star or the cool secondary star. A small third-light corresponding to about 5 % and 2 % of the total systemic light in the B and V bandpasses, respectively, supports the third-body hypothesis proposed by the period study. Preliminary absolute dimensions of the system were derived and used to look into its evolutionary status with other W UMa binaries in the mass-radius and mass-luminosity diagrams. A possible evolution scenario of the system was also discussed in the context of the mass vs mass ratio diagram.

• Korean Chemical Engineering Research
• /
• v.43 no.2
• /
• pp.236-242
• /
• 2005

Membrane fouling by protein mixtures during microfiltration has been investigated for binary mixtures of bovine serum albumin (BSA), casein, lysozyme, pepsin, and ovalbumin. Filtration experiments were carried out using $0.2{\mu}m$ polycarbonate track-etched (PCTE) membrane in a stirred cell under constant transmembrane pressure (14 kPa) and concentration of hydrogen ion (pH=11) to study the effect of mixture composition on filtrate flux decline. Flux decline data were analyzed using a pore blockage-cake formation model developed recently. It was found that the model is in a good agreement with the experimental data. Fouling parameters such as the rate of pore blockage(${\alpha}$), the initial resistance of the protein deposit ($R_{po}$) and the increasing rate of the protein layer resistance(${\beta}$) were used to evaluate the rate of filtrate flow by membrane fouling in the binary mixture system. Generally, the trend of ${\alpha}$ is comparable with that of filtrate flux decline. It was also found that fast flux decreasing was observed over the binary mixture containing casein. The result is due to high value of the initial resistance of the protein deposit ($R_{po}$) over casein.

• Journal of the Korean Data and Information Science Society
• /
• v.27 no.2
• /
• pp.549-557
• /
• 2016

In this paper we propose multiclassification method for large data sets by ensembling least squares support vector machines (LS-SVM) with principal components instead of raw input vector. We use the revised one-vs-all method for multiclassification, which is one of voting scheme based on combining several binary classifications. The revised one-vs-all method is performed by using the hat matrix of LS-SVM ensemble, which is obtained by ensembling LS-SVMs trained using each random sample from the whole large training data. The leave-one-out cross validation (CV) function is used for the optimal values of hyper-parameters which affect the performance of multiclass LS-SVM ensemble. We present the generalized cross validation function to reduce computational burden of leave-one-out CV functions. Experimental results from real data sets are then obtained to illustrate the performance of the proposed multiclass LS-SVM ensemble.

• Journal of Intelligence and Information Systems
• /
• v.23 no.1
• /
• pp.95-108
• /
• 2017

Recently, AlphaGo which is Bakuk (Go) artificial intelligence program by Google DeepMind, had a huge victory against Lee Sedol. Many people thought that machines would not be able to win a man in Go games because the number of paths to make a one move is more than the number of atoms in the universe unlike chess, but the result was the opposite to what people predicted. After the match, artificial intelligence technology was focused as a core technology of the fourth industrial revolution and attracted attentions from various application domains. Especially, deep learning technique have been attracted as a core artificial intelligence technology used in the AlphaGo algorithm. The deep learning technique is already being applied to many problems. Especially, it shows good performance in image recognition field. In addition, it shows good performance in high dimensional data area such as voice, image and natural language, which was difficult to get good performance using existing machine learning techniques. However, in contrast, it is difficult to find deep leaning researches on traditional business data and structured data analysis. In this study, we tried to find out whether the deep learning techniques have been studied so far can be used not only for the recognition of high dimensional data but also for the binary classification problem of traditional business data analysis such as customer churn analysis, marketing response prediction, and default prediction. And we compare the performance of the deep learning techniques with that of traditional artificial neural network models. The experimental data in the paper is the telemarketing response data of a bank in Portugal. It has input variables such as age, occupation, loan status, and the number of previous telemarketing and has a binary target variable that records whether the customer intends to open an account or not. In this study, to evaluate the possibility of utilization of deep learning algorithms and techniques in binary classification problem, we compared the performance of various models using CNN, LSTM algorithm and dropout, which are widely used algorithms and techniques in deep learning, with that of MLP models which is a traditional artificial neural network model. However, since all the network design alternatives can not be tested due to the nature of the artificial neural network, the experiment was conducted based on restricted settings on the number of hidden layers, the number of neurons in the hidden layer, the number of output data (filters), and the application conditions of the dropout technique. The F1 Score was used to evaluate the performance of models to show how well the models work to classify the interesting class instead of the overall accuracy. The detail methods for applying each deep learning technique in the experiment is as follows. The CNN algorithm is a method that reads adjacent values from a specific value and recognizes the features, but it does not matter how close the distance of each business data field is because each field is usually independent. In this experiment, we set the filter size of the CNN algorithm as the number of fields to learn the whole characteristics of the data at once, and added a hidden layer to make decision based on the additional features. For the model having two LSTM layers, the input direction of the second layer is put in reversed position with first layer in order to reduce the influence from the position of each field. In the case of the dropout technique, we set the neurons to disappear with a probability of 0.5 for each hidden layer. The experimental results show that the predicted model with the highest F1 score was the CNN model using the dropout technique, and the next best model was the MLP model with two hidden layers using the dropout technique. In this study, we were able to get some findings as the experiment had proceeded. First, models using dropout techniques have a slightly more conservative prediction than those without dropout techniques, and it generally shows better performance in classification. Second, CNN models show better classification performance than MLP models. This is interesting because it has shown good performance in binary classification problems which it rarely have been applied to, as well as in the fields where it's effectiveness has been proven. Third, the LSTM algorithm seems to be unsuitable for binary classification problems because the training time is too long compared to the performance improvement. From these results, we can confirm that some of the deep learning algorithms can be applied to solve business binary classification problems.

• Journal of Fisheries and Marine Sciences Education
• /
• v.14 no.2
• /
• pp.177-190
• /
• 2002

An experimental study was carried out to make clear heat transfer characteristics in flow boiling of binary mixtures of refrigerants R134a and R123 in a uniformly heated horizontal tube. Experiments were run at a pressure of 0.6 MPa both for pure fluids and mixtures in the ranges of heat flux $10{\sim}50{kW/m}^2$, vapor quality 0~100% and mass flux 150-600 $kg/m^2s$. Heat transfer coefficients of mixtures were reduced compared to the interpolated values between pure fluids both in the low quality region where the nucleate boiling is dominant and in the high quality region where the convective evaporation is dominant. Total pressure drop during two-phase flow boiling in a horizontal tube consists of the sum of two components, that is, the frictional pressure drop and pressure drop due to acceleration. The frictional pressure drop is the most difficult component to predict, and makes the most important contribution to the total pressure drop. On the other hand, the acceleration pressure drop resulting from the variation of the momentum flux caused by phase change is generally small as compared to the frictional pressure drop. There is no significant difference in measured pressure drop between mixtures and pure fluids. The correlation of Martinelli and Nelson predicted most of the present data both for pure and mixed refrigerants within 30%.

• Journal of the Economic Geographical Society of Korea
• /
• v.16 no.1
• /
• pp.1-16
• /
• 2013

This study investigates the spatial distribution characteristics of Korean fashion industries during the last decade, in which the economic geography of fashion industries has changed dynamically with economic globalization and "thus resulted in increased" demand "of" diversification. In particular, this study examines the spatial distribution patterns of fashion industries in the Seoul metropolitan area where fashion industries are highly agglomerated. For the purpose, this study applies Moran's I Index of spatial autocorrelation analysis for seven functional sectors of fashion industries related to fashion production. The global and local agglomeration patterns are examined for each functional sector. The results clarify the distinction in the spatial agglomeration patterns among the seven functional sectors of fashion industries in the Seoul Metropolitan area. Logit models are developed to examine the interrelationships among functional sectors in their spatial agglomeration distribution patterns. By conducting binary logistic regression analysis, we find out how the spatial agglomeration of each functional sector is related to the others.

• Korean Journal of Materials Research
• /
• v.23 no.3
• /
• pp.183-189
• /
• 2013

$Cu_2ZnSn(S,Se)_4$ material is receiving an increased amount of attention for solar cell applications as an absorber layer because it consists of inexpensive and abundant materials (Zn and Sn) instead of the expensive and rare materials (In and Ga) in $Cu(In,Ga)Se_2$ solar cells. We were able to achieve a cell conversion efficiency to 4.7% by the selenization of a stacked metal precursor with the Cu/(Zn + Sn)/Mo/glass structure. However, the selenization of the metal precursor results in large voids at the absorber/Mo interface because metals diffuse out through the top CZTSe layer. To avoid the voids at the absorber/Mo interface, binary selenide compounds of ZnSe and $SnSe_2$ were employed as a precursor instead of Zn and Sn metals. It was found that the precursor with Cu/$SnSe_2$/ZnSe stack provided a uniform film with larger grains compared to that with $Cu_2Se/SnSe_2$/ZnSe stack. Also, voids were not observed at the $Cu_2ZnSnSe_4$/Mo interface. A severe loss of Sn was observed after a high-temperature annealing process, suggesting that selenization in this case should be performed in a closed system with a uniform temperature in a $SnSe_2$ environment. However, in the experiments, Cu top-layer stack had more of an effect on reducing Sn loss compared to $Cu_2Se$ top-layer stack.

• Journal of the Korean Institute of Telematics and Electronics
• /
• v.18 no.5
• /
• pp.1-9
• /
• 1981

Experimental results on a 4-bit charge-coupled A/D converter are described. Major operations in the successive approximation algorithm are implemented in a monolithic chip, CCADC, which was fabricated usir p-channel CCD technology, with its die size of 4,200 mil2 Typical operating frequency range has been found out to be from 500Hz to 200kHz. In that frequency range, no missing code has been found in the whole signal range of 2.4 volts for ramp signal slewing at 1 LSB/(sampling time). A discussion is made on several layout techniques to conserve the nominal binary ratio of (8:4:2:1) among the areas of four adjacent potential wells (M wells), whose charge storing capacities correspond to each bit magnitude - 3.6 pC, 1.8 pC, 0.9 pC, and 0.45 pC nominal in the order of MSB to the LSB. The effect of 'dump slot', which is responsible for the excessive nonlinearity (2$\frac{1}{2}$LSB) in the A/D converter, is explained. A novel input scheme called 'slot zero insertion' to circumvent the deleterious effects of the dump slot is described with the experimental results.