• Biomolecules & Therapeutics
• /
• v.4 no.3
• /
• pp.232-238
• /
• 1996

The physical dependence potency of DA-5018, a non-narcotic analgesic agent, was tested in mice dosed with 0.5 and 4 mg/kg/day for 2 months and daily increasing doses of 1, 2, 4, 6, 8 and 10 mg/kg over 10 days. Physical dependence was assessed taking natural withdrawal induced morphine-type abstinence (jumping, falling, biting or backward locomotion, rearing etc.) as well as barbiturates-type abstinence (body weight reduction, convulsion, ataxia etc.) into consideration. The results were compared with those after the same daily increasing doses of morphine. DA-5018 did not show evidence of physical dependence liability or abuse potential as measured by morphine-type or barbiturate-type abstinence signs following daily increasing or 2-month repeated administration. On the other hand, daily increasing doses of morphine produced physical dependence and the dependent state disappeared about 6 hours after the start of withdrawal signs. In the single dose suppression test, a single dose of morphine completely suppressed natural withdrawal signs that appeared in morphine-dependent animals. Therefore, these results indicate that DA-5018 does not have abuse potential and physical dependence liability.

• Structural Engineering and Mechanics
• /
• v.62 no.2
• /
• pp.171-178
• /
• 2017

Nonlinear vibrations of an Euler-Bernoulli beam resting on a nonlinear elastic foundation are discussed. In search of approximate analytical solutions, the classical multiple scales (MS) and the multiple scales Lindstedt Poincare (MSLP) methods are used. The case of primary resonance is investigated. Amplitude and phase modulation equations are obtained. Steady state solutions are considered. Frequency response curves obtained by both methods are contrasted with each other with respect to the effect of various physical parameters. For weakly nonlinear systems, MS and MSLP solutions are in good agreement. For strong hardening nonlinearities, MSLP solutions exhibit the usual jump phenomena whereas MS solutions are not reliable producing backward curves which are unphysical.

• Proceedings of the Optical Society of Korea Conference
• /
• 2003.07a
• /
• pp.20-21
• /
• 2003

Erbium doped fiber (EDF) sources are useful devices for characterization of optical components for wavelength division multiplexing (WDM) fiber optic communication system. Therefore, there are many efforts to extend the bandwidth and to increase the power of the light source. Especially, L-band ASE source uses the low inversion state of EDF. It makes the power efficiency very low and needs a lot of fiber as several times as the fiber needed in C-band ASE generation. (omitted)

• KDI Journal of Economic Policy
• /
• v.37 no.sup
• /
• pp.33-63
• /
• 2015

This paper attempts to investigate the Korean households' inflation expectations with particular attention to information rigidity. For this purpose, we derive an empirical model from a sticky information model $\acute{a}$ la Mankiw and Reis (2002) and estimate it. In addition, it is also examined whether the expectation formation is state-dependent on macroeconomic conditions. The main findings of this paper are as follows. First, it turns out that the information rigidity in Korean households' inflation expectations is very high. In a month, most of the households simply keep their inflation expectations the same as before instead of updating them based on newly arrived information. Furthermore, when updating their expectations, the households tend to rely on the backward-looking information such as actual inflation rates in the past rather than on the forward-looking forecasts by experts. Second, it is found that the expectation formation is varying as inflation rate changes. Specifically, when the inflation is high, the sensitivity of the households' inflation expectations to actual inflation increases and the gap between inflation expectations and actual inflation shrinks. It implies that Korean households update their expectations more frequently when the inflation matters than not.

• Journal of the Korean Home Economics Association
• /
• v.23 no.3
• /
• pp.27-38
• /
• 1985

In KOREA, the history of weaving is so long even in prinitive, there are the traces of fabric emmision using Spindle already in the Neolithic age. And coming up to the period of Three States, becoming active of cultivation of fibers, it can be known the variety of kinds and the production of fine products by improvement of new method of weaving using weaving machine. In the period of shilla state, there is a record of making the fine fabrics with 28 bracts. But wearing of silks fabrics was limitted for only the nobility and common people were wearing native thick hemp clothes. Also in the period of Korea state, they were wearing the hemp clothes by cultivation of hemp. The good quality products were worn by the King or the nobility and women in KOREA made their best with whole efforts to pay to the authorities with woven products of hemp clothes, so in the 14th year of King Chung- Ryul, King prohibited the presentation of fine hemp clothes by his order. By the end of KOREA state, before the production, common people was mainly wearing the hemp clothes. Coming up to the Kingdom of Chosun the sericulture was promoted by establishment of the sericulture encouraging low. Therefore the working hours of women were highly increased. The products of Song-do, Chin-ju for cotton clothes, those of Han-san-the same now as in old times-for ramie clothes, those of Han-Kyung province and An-dong for hemp fabrics were estimated as the best qualities. And the hemp clothes of Kang-won province is not so fine but is very useful and famous for mourning clothes and summer clothes for the farmer. It is true that our history of weaving was begun with hemp as a continous and precious friend of common people during all the historical periods-even though for a while it went backward because of chinese silks.

• Journal of Intelligence and Information Systems
• /
• v.22 no.2
• /
• pp.127-142
• /
• 2016

Deep learning model is a kind of neural networks that allows multiple hidden layers. There are various deep learning architectures such as convolutional neural networks, deep belief networks and recurrent neural networks. Those have been applied to fields like computer vision, automatic speech recognition, natural language processing, audio recognition and bioinformatics where they have been shown to produce state-of-the-art results on various tasks. Among those architectures, convolutional neural networks and recurrent neural networks are classified as the supervised learning model. And in recent years, those supervised learning models have gained more popularity than unsupervised learning models such as deep belief networks, because supervised learning models have shown fashionable applications in such fields mentioned above. Deep learning models can be trained with backpropagation algorithm. Backpropagation is an abbreviation for "backward propagation of errors" and a common method of training artificial neural networks used in conjunction with an optimization method such as gradient descent. The method calculates the gradient of an error function with respect to all the weights in the network. The gradient is fed to the optimization method which in turn uses it to update the weights, in an attempt to minimize the error function. Convolutional neural networks use a special architecture which is particularly well-adapted to classify images. Using this architecture makes convolutional networks fast to train. This, in turn, helps us train deep, muti-layer networks, which are very good at classifying images. These days, deep convolutional networks are used in most neural networks for image recognition. Convolutional neural networks use three basic ideas: local receptive fields, shared weights, and pooling. By local receptive fields, we mean that each neuron in the first(or any) hidden layer will be connected to a small region of the input(or previous layer's) neurons. Shared weights mean that we're going to use the same weights and bias for each of the local receptive field. This means that all the neurons in the hidden layer detect exactly the same feature, just at different locations in the input image. In addition to the convolutional layers just described, convolutional neural networks also contain pooling layers. Pooling layers are usually used immediately after convolutional layers. What the pooling layers do is to simplify the information in the output from the convolutional layer. Recent convolutional network architectures have 10 to 20 hidden layers and billions of connections between units. Training deep learning networks has taken weeks several years ago, but thanks to progress in GPU and algorithm enhancement, training time has reduced to several hours. Neural networks with time-varying behavior are known as recurrent neural networks or RNNs. A recurrent neural network is a class of artificial neural network where connections between units form a directed cycle. This creates an internal state of the network which allows it to exhibit dynamic temporal behavior. Unlike feedforward neural networks, RNNs can use their internal memory to process arbitrary sequences of inputs. Early RNN models turned out to be very difficult to train, harder even than deep feedforward networks. The reason is the unstable gradient problem such as vanishing gradient and exploding gradient. The gradient can get smaller and smaller as it is propagated back through layers. This makes learning in early layers extremely slow. The problem actually gets worse in RNNs, since gradients aren't just propagated backward through layers, they're propagated backward through time. If the network runs for a long time, that can make the gradient extremely unstable and hard to learn from. It has been possible to incorporate an idea known as long short-term memory units (LSTMs) into RNNs. LSTMs make it much easier to get good results when training RNNs, and many recent papers make use of LSTMs or related ideas.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.25 no.4A
• /
• pp.546-552
• /
• 2000

The decoding of Turbo-Code relies on the application of a soft input/soft output decoders which can be realized using maximum-a-posteriori(MAP) symbol estimator[l]. Radix-2 MAP decoder can not be used for high speed communications because of a large number of interleaver block size N. This paper proposed a new simple method for radix-4 MAP decoder based on radix-2 MAP decoder in order to reduce the interleave block size. A branch metrics, forward and backward recursive functions are proposed for applying to radix-4 MAP structure with symbol interleaver. Radix-4 MAP decoder shall be illustratively described and its error performance capability shall be compared to conventional radix-2 MAP decoder in AWGN channel.

• The Journal of Korea Robotics Society
• /
• v.6 no.4
• /
• pp.323-333
• /
• 2011

Due to the limited pendulum motion range, the conventional one-pendulum driven spherical robot has limited driving capability. Especially it can not drive parallel direction with center horizontal axis to which pendulum is attached from stationary state. To overcome the limited driving capability of one-pendulum driven spherical robot, we introduce a spherical robot, called KisBot II, with a new type of curved two-pendulum driving mechanism. A cross-shape frame of the robot is located horizontally in the center of the robot. The main axis of the frame is connected to the outer shell, and each curved pendulum is connected to the end of the other axis of the frame respectively. The main axis and pendulums can rotate 360 degrees inside the sphere orthogonally without interfering with each other, also the two pendulums can rotate identically or independent of each other. Due to this driving mechanism, KisBot II has various motion generation abilities, including a fast steering, turning capability in place and during travelling, and four directions including forward, backward, left, and right from stationary status. Experiments for several motions verify the driving efficiency of the proposed spherical robot.

• Journal of the Korean Home Economics Association
• /
• v.49 no.1
• /
• pp.1-16
• /
• 2011

This study analyzed the items for the Home Economics teacher selection test, which were carried out through the school year 2002 to 2010, based on Pedagogical Content Knowledge in Home Economics Education (H-PCK). Two frameworks were used: (1) H-PCK matrix transformed from the Content Knowledge in Home Economics Education (H-CK) and the Pedagogical Knowledge; (2) Five components of H-PCK. The results of this study were as follows: (1) The test items were classified into H-CK items and H-PCK items from the H-PCK matrix analysis. (2) While the number of items on knowledge of teaching strategies was the highest, no item on knowledge of understanding students was identified. Since the test can be considered as Measurement-Driven Instruction, it will have backward influences on the teacher education program in universities. Thus, if promising and capable new Home Economics teachers are to be recruited, the test items should include all the components of H-PCK.

• Proceedings of the KSME Conference
• /
• 2004.11a
• /
• pp.1122-1127
• /
• 2004

an isolated droplet combustion exposed to pressure perturbations in stagnant gaseous environment is numerically conducted. Governing equations are solved for flow parameters at gas and liquid phases separately and thermodynamic parameters at the interfacial boundary are matched for problem closure. For high-pressure effects, vapor-liquid interfacial thermodynamics is rigorously treated. A series of parametric calculations in terms of mean pressure level and wave frequencies are carried out employing a n-pentane droplet in stagnant gaseous air. Results show that the operating pressure and driving frequency have an important role in determining the amplitude and phase lag of a combustion response. Mass evaporation rate responding to pressure waves is amplified with increase in pressure due to substantial reduction in latent heat of vaporization. Phase difference between pressure and evaporation rate decreases due to the reduced thermal inertia at high pressure. In addition to this, augmentation of perturbation frequency also enhances amplification of vaporization rate because the time period for the pressure oscillation is much smaller than the liquid thermal inertia time. The phase of evaporation rate shifts backward due to the elevated thermal inertia at high acoustic frequency.