• Journal of Intelligence and Information Systems
• /
• v.25 no.3
• /
• pp.43-62
• /
• 2019

At one time, the anomaly detection sector dominated the method of determining whether there was an abnormality based on the statistics derived from specific data. This methodology was possible because the dimension of the data was simple in the past, so the classical statistical method could work effectively. However, as the characteristics of data have changed complexly in the era of big data, it has become more difficult to accurately analyze and predict the data that occurs throughout the industry in the conventional way. Therefore, SVM and Decision Tree based supervised learning algorithms were used. However, there is peculiarity that supervised learning based model can only accurately predict the test data, when the number of classes is equal to the number of normal classes and most of the data generated in the industry has unbalanced data class. Therefore, the predicted results are not always valid when supervised learning model is applied. In order to overcome these drawbacks, many studies now use the unsupervised learning-based model that is not influenced by class distribution, such as autoencoder or generative adversarial networks. In this paper, we propose a method to detect anomalies using generative adversarial networks. AnoGAN, introduced in the study of Thomas et al (2017), is a classification model that performs abnormal detection of medical images. It was composed of a Convolution Neural Net and was used in the field of detection. On the other hand, sequencing data abnormality detection using generative adversarial network is a lack of research papers compared to image data. Of course, in Li et al (2018), a study by Li et al (LSTM), a type of recurrent neural network, has proposed a model to classify the abnormities of numerical sequence data, but it has not been used for categorical sequence data, as well as feature matching method applied by salans et al.(2016). So it suggests that there are a number of studies to be tried on in the ideal classification of sequence data through a generative adversarial Network. In order to learn the sequence data, the structure of the generative adversarial networks is composed of LSTM, and the 2 stacked-LSTM of the generator is composed of 32-dim hidden unit layers and 64-dim hidden unit layers. The LSTM of the discriminator consists of 64-dim hidden unit layer were used. In the process of deriving abnormal scores from existing paper of Anomaly Detection for Sequence data, entropy values of probability of actual data are used in the process of deriving abnormal scores. but in this paper, as mentioned earlier, abnormal scores have been derived by using feature matching techniques. In addition, the process of optimizing latent variables was designed with LSTM to improve model performance. The modified form of generative adversarial model was more accurate in all experiments than the autoencoder in terms of precision and was approximately 7% higher in accuracy. In terms of Robustness, Generative adversarial networks also performed better than autoencoder. Because generative adversarial networks can learn data distribution from real categorical sequence data, Unaffected by a single normal data. But autoencoder is not. Result of Robustness test showed that he accuracy of the autocoder was 92%, the accuracy of the hostile neural network was 96%, and in terms of sensitivity, the autocoder was 40% and the hostile neural network was 51%. In this paper, experiments have also been conducted to show how much performance changes due to differences in the optimization structure of potential variables. As a result, the level of 1% was improved in terms of sensitivity. These results suggest that it presented a new perspective on optimizing latent variable that were relatively insignificant.

• Journal of Life Science
• /
• v.29 no.9
• /
• pp.1002-1009
• /
• 2019

Early life stress (ELS) increases the risk of depression. ELS may be involved in the susceptibility to subsequent stress exposure during adulthood. We investigated whether epigenetic mechanisms of p11 promoter affect the vulnerability to chronic unpredictable stress (CUS) induced by the maternal separation (MS). Mice pups were separated from their dams (3 hr/day from P1-P21). When the pups reached adulthood, we applied CUS (daily for 3 weeks). The levels of hippocampal p11 expression were analyzed by quantitative real-time PCR. The levels of acetylated and methylated histone H3 at p11 promoter were measured by chromatin immunoprecipitation. Depression-like behavior was measured by the forced swimming test (FST). The MS and CUS group exhibited significant decreases in p11 mRNA level and the MS plus CUS group had a greater reduction in this level than the CUS group. The MS plus CUS group also resulted in greater reduction in H3 acetylation than the CUS group. This reduction was associated with an upregulation of histone deacetylase 5. Additionally, the MS plus CUS group showed a greater decrease in H3K4met3 level and a greater increase in H3K27 met3 level than the CUS group. Consistent with the reduction of p11 expression, the MS plus CUS group displayed longer immobility times in the FST compared to the control group. Mice exposed to MS followed by CUS had much greater epigenetic alterations in the hippocampus compared to adult mice that only experienced CUS. ELS can exacerbate the effect of stress exposure during adulthood through histone modification of p11 gene.

• The Journal of the Petrological Society of Korea
• /
• v.28 no.3
• /
• pp.157-169
• /
• 2019

Pseudotachylytes, produced by frictional heating during seismic slip, provide information that is critical to understanding the physics of earthquakes. We report the results of occurrence, structural characteristics, scanning electron microscopic observation and geochemical analysis of pseudotachylytes, which is presumed to have formed after the Late Cretaceous in outcrops of the Paleoproterozoic granitic gneiss on the Bulil waterfall of the Jirisan area, Yeongnam massif, Korea. Fault rocks, which are the products of brittle deformation under the same shear stress regime in the study area, are classified as pseudotachylyte and foliated cataclasite. The occurrences of pseudotachylyte identified on the basis of thickness and morphology are fault vein-type and injection vein-type pseudotachylyte. A number of fault vein-type pseudotachylytes occur as thin (as thick as 2 cm) layers generated on the fault plane, and are cutting general foliation and sheared foliation developed in granitic gneiss. Smaller injection vein-type pseudotachylytes are found along the fault vein-type pseudotachylytes, and appear in a variety of shapes based on field occurrence and vein geometry. At a first glance fault vein-type seudotachylyte looks like a mafic vein, but it has a chemical composition almost identical to the wall rock of granitic gneiss. Also, it has many subrounded clasts which consist predominantly of quartz, feldspar, biotite and secondary minerals including clay minerals, calcite and glassy materials. Embayed clasts, phenocryst with reaction rim, oxide droplets, amygdules, and flow structures are also observed. All of these evidences indicate the pseudotachylyte formed due to frictional melting of the wall rock minerals during fault slip related to strong seismic faulting events in the shallow depth of low temperature-low pressure. Further studies will be conducted to determine the age and mechanical aspect of the pseudotachylyte formation.

• Korean Journal of Heritage: History & Science
• /
• v.38
• /
• pp.329-358
• /
• 2005

The Three storied Stone pagoda in Gameunsa Temple site, one of the early staged stone pagodas, has been known as a standard for Silla stone pagodas. A stone pagoda is not only a stone art work and but also a stone structure. Most studies and investigation of the stone pagoda has done mainly based on style and chronological research according to an art historical view. However, there is not an attempt to research the stone pagoda as a stone architecture. Most Korean experts at the stone pagoda has art history in their background. Engineers who can understand the structure of the stone pagoda are very limited. More architectural and engineering approach is need to research not only art historial understanding but also safety as a structure. We can find many technical know-how from our ancestors who made stone pagodas. 1. To reduce any deformation such as relaxation and sinking of BuJae which is caused by a heavy load, the BuJae (consist of a foundation stone and lower stereobates) should be enlarged. 2. A special construction method for connection between Myonsuk and Tangjoo was invented. This unique method is not used any longer after the Three storied Stone pagoda in Gameunsa Temple site. 3. The upper BuJae and the lower BuJae are missed each other by making a difference of Okgaesuk and Okgaebatchim in size. It is done for a distribution of perpendicular load and a prevention for relaxation of BuJae. 4. The center of gravity in the BuJae is located to the center of the stone pagoda by trimming the upper surface of the Okgaebatchim into a convex shape. The man who made stone pagodas had excellent knowledge on the engineering and techniques to understand the structure of the stone pagodas. We can confirm it as follows: the enlarged BuJae, dislocated connection between upper Bujae and lower BuJae, and moving the center of gravity close to the center of the stone pagoda.

• Korean Journal of Heritage: History & Science
• /
• v.40
• /
• pp.243-270
• /
• 2007

Linked pearls pattern expressed on textiles have no limited scale or shape when manufacturing, so they are free in expression. And from the design, material, and color we can analogize the social culture of that age. Oriental linked pearls pattern was started from the Sasanian Persia and introduced through the Silk Road, so it is closely connected with the East and the West culture. This study will consider from the 5th century to the 10th century; the mural costume of the West Central Asia, the ancient textiles excavated from the Sinjiang and Qinghai area of China, and the linked pearls pattern which are collected at Shosoin, Japan. And from this study, will concentrate on clarifying the linked pearls pattern's condition of the cultural exchange between the East and the West and it's structural variation process. The design of linked pearls pattern delivered to the East through the Silk Road is differed by area. For example, in the Western Pamir Plateau, where the ancient Sogdians mainly lived, the excavated linked pearls pattern's subject were deer or cassowary variated from the West Asian motif. But the ones excavated from Kucha Xingang had Chinese motifs added so they showed Chinese characters or Buddhist Bodhisattva image instead of Helios. Like this, the appearance of new patterns, which were accompanied by structural variations, gradually deviated from the standardized pattern of the Sasanian Persia. And this structural variation process has relations with the construction and arrangement method of various patterns of the after ages. The foliated floral Spray, which is placed at the lozenge space of linked pearls' space, had developed into ogival - shaped pattern (Neunghwamun). And the prevalence of geometrical structure pattern after the 10th century and the unfolding method of Tapjamun which is arranging unit pattern in order, are similar to the linked pearl pattern. In brief, linked pearls pattern accompanied by technical improvement let us understand the polished artistic code from its expression, and has importance in showing universal pattern beyond region and culture.

• The Sea:JOURNAL OF THE KOREAN SOCIETY OF OCEANOGRAPHY
• /
• v.24 no.2
• /
• pp.332-350
• /
• 2019

Zonal hydrographic section measurements at $39.7^{\circ}N$ were conducted between $129.0^{\circ}E$ and $131.3^{\circ}E$ from August 7 to 25 in 2017 using an underwater glider. The glider traveled about 440 km for about 18 days along the 106 line of the regular shipboard measurements in the National Institute of Fishery Science (NIFS) and obtained twice a hydrographic section with high horizontal resolution. Even under the strong East Korea Warm Current with maximum speed of 0.8 m/s across the section, the glider successfully maintained the designated path within an RMS distance of 400 m. By comparing with the NIFS shipboard hydrographic section, it is confirmed that high spatial resolution measurements obtained from a glider were necessary to properly observe front and eddy variability in the East Sea where a typical spatial scale is smaller than the open oceans. From the glider section measurements, a new lens-shaped eddy was found in the thermocline. The lens-shaped anticyclonic eddy had 10~13 km in horizonal width and about 200 m in height like a typical submesoscale eddy resided within the thermocline, which was firstly named as Korea intrathermocline eddy (Keddy). The Keddy has the distinguishing characteristics of a typical intrathermocline eddy, such as a central core with anomalously weak stratification, a convex shaped lens bounded by the stratification anomaly, an interior maximum of velocity at 170 m, no surface appearance of the geopotential field, a small or comparable horizontal width relative to the first baroclinic Rossby radius of deformation, and the Rossby nubmer of 0.7.

• Journal of the Korean Institute of Landscape Architecture
• /
• v.47 no.1
• /
• pp.26-38
• /
• 2019

This study is based on the analysis of the characteristics of waterfront scenery. Recently, waterfront development has expanded residentially, commercially and into leisure space. In the development of the waterfront, it is necessary to apply designs suitable for urban and various other waterfront areas. In this study, the natural scenery of the waterfront was researched with respect to the Korean Jingyeong landscape paintings and the main elements of the scenery were analyzed. In this study, 105 painting of Korean Jingyeong landscapes paintings were selected for the analysis of the waterside scenery. The paintings of Jeong Seon were studied to categorize streams topographically into mountainous, upper, middle, lower, and ocean types. In addition, major micro-topography elements, which are 13 water image elements and 13 staffage elements were analyzed. The main waterfront landscape elements are divided into 13 types. The waterfalls were divided into long waterfalls, short waterfalls, cascading waterfalls, and other aspects considered were line stream, curve stream, multi-curve stream, pond, water surface, flow surface, wave surface, rock side, pile sandy side, sandy side. There are 13 kinds of staffage elements, include pine forest, pine trees, fir trees, bamboo trees, willow trees, broadleaf tree, villages, houses, gazebo, boat, bridges, and people. The waterfront landscape by a river area was explained according to each characteristic of the waterfront landscape and staffage, and their changes were analyzed in each area. The 105 paintings were divided into 35 pieces of mountainous streams, 9 upper streams, 5 middle streams, 35 lower streams, and 21 oceans, and the change of each waterfront landscape and staffage was analyzed. Based on the topographical analysis of the waterfront landscape and staffage, the results can be summarized into 5 types of the waterfront landscape. Based on the micro-topographical characteristics of the waterfront landscape styles are as follow. In the mountainous streams, long waterfall and deep forest type are apparent, which depicts deep mountain waterfall scenery, and a multi-stream forest is the scenery of a picnic in the mountains, which is a representative form of mountainous streams landscape. In the upper-middle stream, the water-surface and gazebo type is predominant. In the lower stream, the sandy-gazebo typ scenery is predominant and the sandy depiction is unique to lower stream landscape. Pile sandy-dock type is life scenes where human activity highlighted, is a representative form of the lower stream landscapes. The characteristic of the coastal landscape is the serpentine rock scenery on the beach and the wave-serpentine rock type that forms the main coastal landscape. The study aims to propose significant design elements for a natural waterfront landscape planning based on the analysis of landscape in the paintings of Jeong Seon.

• Journal of Bio-Environment Control
• /
• v.28 no.2
• /
• pp.95-103
• /
• 2019

Temperature and relative humidity are important factors in crop cultivation and should be properly controlled for improving crop yield and quality. In order to control the environment accurately, we need to predict how the environment will change in the future. The objective of this study was to predict air temperature and relative humidity at a future time by using a multilayer perceptron (MLP). The data required to train MLP was collected every 10 min from Oct. 1, 2016 to Feb. 28, 2018 in an eight-span greenhouse ($1,032m^2$) cultivating mango (Mangifera indica cv. Irwin). The inputs for the MLP were greenhouse inside and outside environment data, and set-up and operating values of environment control devices. By using these data, the MLP was trained to predict the air temperature and relative humidity at a future time of 10 to 120 min. Considering typical four seasons in Korea, three-day data of the each season were compared as test data. The MLP was optimized with four hidden layers and 128 nodes for air temperature ($R^2=0.988$) and with four hidden layers and 64 nodes for relative humidity ($R^2=0.990$). Due to the characteristics of MLP, the accuracy decreased as the prediction time became longer. However, air temperature and relative humidity were properly predicted regardless of the environmental changes varied from season to season. For specific data such as spray irrigation, however, the numbers of trained data were too small, resulting in poor predictive accuracy. In this study, air temperature and relative humidity were appropriately predicted through optimization of MLP, but were limited to the experimental greenhouse. Therefore, it is necessary to collect more data from greenhouses at various places and modify the structure of neural network for generalization.

• Journal of Korean Society of Coastal and Ocean Engineers
• /
• v.30 no.6
• /
• pp.306-318
• /
• 2018

In order to evaluate the seismic capacity of massive vertical type breakwaters which have intensively been deployed along the coast of South Korea over the last two decades, we carry out the preliminary numerical simulation against the PoHang, GyeongJu, Hachinohe 1, Hachinohe 2, Ofunato, and artificial seismic waves based on the measured time series of ground acceleration. Numerical result shows that significant sliding can be resulted in once non-negligible portion of seismic energy is shifted toward the longer period during its propagation process toward the ground surface in a form of shear wave. It is well known that during these propagation process, shear waves due to the seismic activity would be amplified, and non-negligible portion of seismic energy be shifted toward the longer period. Among these, the shift of seismic energy toward the longer period is induced by the viscosity and internal friction intrinsic in the soil. On the other hand, the amplification of shear waves can be attributed to the fact that the shear modulus is getting smaller toward the ground surface following the descending effective stress toward the ground surface. And the weakened intensity of soil as the number of attacking shear waves are accumulated can also contribute these phenomenon (Das, 1993). In this rationale, we constitute the numerical model using the model by Hardin and Drnevich (1972) for the weakened shear modulus as shear waves go on, and shear wave equation, in the numerical integration of which $Newmark-{\beta}$ method and Modified Newton-Raphson method are evoked to take nonlinear stress-strain relationship into account. It is shown that the numerical model proposed in this study could duplicate the well known features of seismic shear waves such as that a great deal of probability mass is shifted toward the larger amplitude and longer period when shear waves propagate toward the ground surface.

• Korean Journal of Breeding Science
• /
• v.50 no.4
• /
• pp.406-414
• /
• 2018

This study was conducted to develop environmental risk assessments and biosafety guides for insect-resistant genetically modified rice in an LMO (Living Modified Organism) isolation field. In the LMO quarantine area of Kyungpook National University, the species diversities and population densities of non-target insects found on insect-resistant genetically modified rice (Bt-T), rice resistant to Cnaphalocrocis medinalis, and non-GM rice (Dongjin-byeo and Ilmi-byeo) were investigated. The Bt-T plants were, therefore, evaluated under field conditions to detect possible impacts on above ground insects and spiders. In 2016 and 2017, the study compared transgenic rice and two non-GM reference rice, namely Dongjin-byeo and Ilmi-byeo, at Gunwi. A total of 9,552 individuals from 51 families and 11 orders were collected from the LMO isolation field. From the three types of rice fields, a total of 3,042; 3,212; and 3,297 individuals from the Bt-T, Dongjin-byeo, and Ilmi-byeo were collected, respectively. There was no difference between the population densities of the non-target insect pests, natural enemies, and other insects on the Bt-T compared to non-GM rice. The data on insect species population densities were subjected to principal component analysis (PCA) without distinguishing between the three varieties, namely GM, non-GM, and reference cultivar, in all cultivation years. However, the PCA clearly separated the samples based on the cultivation years. These results suggest that insect species diversities and population densities during plant cultivation are determined by environmental factors (growing condition and seasons) rather than by genetic factors.