• 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.

• Journal of Korean Classical Literature and Education
• /
• no.38
• /
• pp.201-238
• /
• 2018

As a way of enhancing the intercultural ability needed for diverse cultural eras, this study focuses on the "narration" of the Italian education scholar Maddalena De Carlo in order to determine the "diverse values" created by the "symbolic representation" based on the folktales narrated by immigrants living in Korea. Through this, it specifically presents educational elements and contents that can raise relative sensitivity. The authors of this paper have connected, empathized, and communicated with people of various cultures in order to go beyond Carlo's discussion. The paper discusses the expansion of cultural sensitivity as an element of education through narrative topics using the folktales of immigrant narrators in Korea. It also recognizes the limitations of a desire for a homogeneous union within an intercultural society and thus formulates educational contents for creating a relationship with heterogeneous ideas through the elimination of communication barriers through heterogeneity and a consideration of the surface and the back. This is systemized in six steps. Step 1: Listening to oral folktales of immigrants, Step 2: Finding heterogeneous motifs imprinted in the immigrants' memories, Step 3: Understanding the meaning of the opposing qualities symbolized by heterogeneous motifs, Step 4: Creating narrative topics containing the key motifs, Step 5: Generating the value of symbolic representation as a narrative topic, and Step 6: Expanding the value of life into a cultural symbol. In Chapter 3, this study focuses on educational contents using immigrants' folktales by applying these six steps. The class contents include the recognition of the limitations of desire for a homogeneous union within an intercultural society and the consideration of how to create a relationship with heterogeneous ideas through the elimination of communication barriers through heterogeneity and consideration of the surface and the back. This paper then compares the Indonesian folktale, The Inverted Ship Mountain and the Mom's Mountain, with the world-famous Oedipus myth, to determine what the symbolic representation of these heterogeneous motifs is. In Step 6, when the symbolic system is culturally extended, the incestuous desire that appears in the "inverted ship" is interpreted as a fixation that was created when the character sought to unite with homogenous idea. The Cambodian folktale, The Girl and the Tiger, is a story that is reminiscent of the Korean folktale, The Old Man with a Lump. Through the motif in "Tiger," this paper generates a narrative topic that will enhance the students' intercultural abilities by culturally expanding their skills in how to relate with a heterogeneous being that is usually represented as an animal. The Vietnamese folktale, The Coconut Bowl, similar to the Korean folktale, GureongDeongDeong SinSeonBi, is a story that draws a variety of considerations about the surface and theback, and it shows readers how to build a relationship with a heterogeneous idea and how to develop and grow with such a relationship. Thus, if a narrative topic is generated and readers are able to empathize using an opposing feature formed by the core motif of the folktale, it becomes possible, through immigrant folklore, to construct a possibility of a new life through the formation of a relationship with an unfamiliar and heterogeneous culture.

• Asian Journal of Turfgrass Science
• /
• v.23 no.2
• /
• pp.219-228
• /
• 2009

The construction procedures and artificial turf maintenance program on golf course definitely influence on the distortion of its environment. Soil microbial communities in soil profile were affected directly by those practises on turf areas. In Jeju island, the environmental impact assessment has been required to apply the first quality class granular activated carbon(GAC), which has a high absorbent character to agricultural chemicals, on the soil profiles of golf green system to reduce the pesticide leaching to ground water. This research was carried out to analyze the changes of microbial communities and chemical properties on soil profiles where GAC had been applied at the construction stage at two golf courses in Jeju. The changes of soil microbial population and chemical properties associated with construction methods of soil profile and agrochemical management program were analyzed by monthly at the surface and sub-soil profiles during April through October, 2007. The total numbers of bacteria and fungi, soil moisture content, soil physio-chemical properties were measured on greens and fairways of the both golf courses with different GAC treatment on the green and fairway soil profiles. The results showed that GAC had positive effects on the water holding capacity, pH and EC, however, it did not improved the holding capacity of available nutrients ${NO_3}^-,{NH_4}^+$, and phosphorus by its sorption phenomenon. In microbial count test, the total numbers of bacteria and fungi showed a great variation during sampling dates. That may directly relate to the agrochemical application, however, the ratio of total bacterial number versus total fungus number showed a constant value on a sub-soil of 15~30cm depth. Thus, the construction method of GAC in soil profile, and application of fertilizer and pesticide, both impacted on the changes of microbial population. It's means that the construction method of soil profile and turf management using agro-materials might greatly affect on the turfgrass culture and the environment of golf course.

• Korean Journal of Remote Sensing
• /
• v.37 no.5_3
• /
• pp.1447-1460
• /
• 2021

Urbanization increases the amount of impervious surface and artificial heat emission, resulting in urban heat island (UHI) effect. Local climate zones (LCZ) are a classification scheme for urban areas considering urban land cover characteristics and the geometry and structure of buildings, which can be used for analyzing urban heat island effect in detail. This study aimed to examine the UHI effect by urban structure in Suwon and Daegu using the LCZ scheme. First, the LCZ maps were generated using Landsat 8 images and convolutional neural network (CNN) deep learning over the two cities. Then, Surface UHI (SUHI), which indicates the land surface temperature (LST) difference between urban and rural areas, was analyzed by LCZ class. The results showed that the overall accuracies of the CNN models for LCZ classification were relatively high 87.9% and 81.7% for Suwon and Daegu, respectively. In general, Daegu had higher LST for all LCZ classes than Suwon. For both cities, LST tended to increase with increasing building density with relatively low building height. For both cities, the intensity of SUHI was very high in summer regardless of LCZ classes and was also relatively high except for a few classes in spring and fall. In winter the SUHI intensity was low, resulting in negative values for many LCZ classes. This implies that UHI is very strong in summer, and some urban areas often are colder than rural areas in winter. The research findings demonstrated the applicability of the LCZ data for SUHI analysis and can provide a basis for establishing timely strategies to respond urban on-going climate change over urban areas.

• Journal of Korean Society of Forest Science
• /
• v.110 no.4
• /
• pp.504-515
• /
• 2021

To examine the relationship between environmental factors influencing the habitat of Aconitum austrokoreense Koidz., this study employed the MexEnt model to evaluate 21 environmental factors. Fourteen environmental factors having an AUC of at least 0.6 were found to be the age of stand, growing stock, altitude, topography, topographic wetness index, solar radiation, soil texture, mean temperature in January, mean temperature in April, mean annual temperature, mean rainfall in January, mean rainfall in August, and mean annual rainfall. Based on the response curves of the 14 descriptive factors, Aconitum austrokoreense Koidz. on the Baekun Mountain were deemed more suitable for sites at an altitude of 600 m or lower, and habitats were not significantly affected by the inclination angle. The preferred conditions were high stand density, sites close to valleys, and distribution in the northwestern direction. Under the five-age class system, the species were more likely to be observed for lower classes. The preferred solar radiation in this study was 1.2 MJ/m². The species were less likely to be observed when the topographic wetness index fell below the reference value of 4.5, and were more likely observed above 7.5 (reference of threshold). Soil analysis showed that Aconitum austrokoreense Koidz. was more likely to thrive in sandy loam than clay. Suitable conditions were a mean January temperature of - 4.4℃ to -2.5℃, mean April temperature of 8.8℃-10.0℃, and mean annual temperature of 9.6℃-11.0℃. Aconitum austrokoreense Koidz. was first observed in sites with a mean annual rainfall of 1,670- 1,720 mm, and a mean August rainfall of at least 350 mm. Therefore, sites with increasing rainfall of up to 390 mm were preferred. The area of potential habitats having distributive significance of 75% or higher was 202 ha, or 1.8% of the area covered in this study.

• Cross-Cultural Studies
• /
• v.42
• /
• pp.7-22
• /
• 2016

Since the Taoyuanming of Confucius an Arcadia (Taohuayuan), utopia imagined in China has continued to this day. Such utopian imagination has been shaped by a variety of literary texts. These works depict the human desires and frustrations that form the utopian imagination. Typically, an Arcadia (Taohuayuan) has been recognized as a symbol of East Asian utopia. An Arcadia (Taohuayuan) anarchism is the utopian character of this small country, based on sensitization of the Lao strong. An Arcadia (Taohuayuan) world ruling class does not exist. But Confucius thought a utopian world is possible to imagine on the premise of 'Virtuous Policy' (德治). By the late Qing Dynasty (淸末), Kangyouwei (康有爲) had written and presented a utopia realized through the system of the "East-West stand." The literary text on utopia information in continued to experiment with implementation of the 'imagined' in the literature as "reality" of everyday life, immediately following the 1911 revolution in the People's Republic of China. As an Avant garde writer, Gefei's "Renmiantaohua" was influenced by the reflexive nature of the report on the point of the experiment. Gefei's "Renmiantaohua" has been rated as an outstanding work depicting the implementation process and the frustration of the utopian imagination. He is an inspiration to vanguard artists who focus on the desire and the frustration of the utopian imagination through the "Renmiantaohua" myths and stories that are comparable to the Sijipuseu story, which in itself perpetuates utopia in literature. In this paper, we explored the trails to implementation of a utopian imagination that has persisted since the ancient Chinese in current literary text. An Arcadia (Taohuayuan) of Confucius and Taoyuanming accompanied Gefei in the 21st century in describing the process of desire and frustration to realize utopia through a variety of traditional shape figures among other favorites. The author interprets those frustrations and desires of the human life course as just utopian imagination.

• Korean Journal of Heritage: History & Science
• /
• v.47 no.3
• /
• pp.42-59
• /
• 2014

Han Pilgyo (1807~1878) was a scholar-official in the later period of the Joseon Dynasty. The research results on spatial structure and landscape techniques of the central government office reviewed through the Sukcheonjeado(宿踐諸衙圖) album collection edited by Han Pilgyo are as follows. First, Sukcheonjeado(宿踐諸衙圖) using Sabangjeondomyobeob(四方顚倒描法, a Korean traditional drawing type) is uniquely proven historical data which helps to understand the spatial structure of the center and local government offices and the characteristics of cultural landscape. Secondly, the central government office located in Yookcho(六曹) Street which is the outside Gwanghwamun(光化門) of the Main Palace(Gyeongbokgung, 景福宮) of the Joseon Dynasty has a center facility(Dangsangdaecheong, 堂上大廳) and attached buildings which are distributed from high to low or from left to right, according to its order of presidency in square-shaped portion of land. The main building was located facing south and by considering the administrative convenience, the environmental effect and the practice of Confucian norms this structure reflects a hierarchical landuse system. Thirdly, the main buildings such as Dangsangdaecheong and Hyangcheong(鄕廳), which are the working place for government officials had large square front yards for constructing a practical patio garden. The back garden was tended to reflect the meaning landscape, with such as pond and pavilion. A particular point was the repeated crossing of active space and passive space(movement and stillness, building and yard, yard and garden), which implements the Yin-Yang principle. Fourth, the characteristics that can be extracted from the central government office landscapes are (1) expandability of outdoor space, connects of front gardens, emphasizes the characteristic of serviceable gardens and back gardens, which in turn emphasizes scenic sides, (2) introduction of water features(square-shaped ponds) that can be used as fire-water and considers environmental-amenity and landscape characteristics, (3) introduction of pavilions for relaxation, mental and physical discipline, and the development of back gardens, (4) significance of Jeongsimsoo(庭心樹) in such things as selection of concise landscape plants like lotus, willow, pine, zelkova and so on, and limited plant introduction, (5) environmental design techniques which set importance on not only aesthetics and ideality but also practical value. Thus, these aspects of the government office landscape can be said to be the universality and particularity of Korean traditional landscape technique and can be extracted similarly in the palaces, temples, lecture halls, and houses of the upper class of the Joseon Dynasty.

• Journal of vocational education research
• /
• v.37 no.3
• /
• pp.85-111
• /
• 2018

The purpose of this study is to derive the competence of nursing teacher in Specialized vocational high school. The knowledge, skills, and attitudes required for the nursing teachers were verified and the competency groups and sub-competencies were verified and priorities were suggested. To do this, 23 nursing teachers who were currently working in the Specialized vocational high school were selected as expert panels and the study was conducted using Delphi and Layered Analysis(AHP) technique. The results of this study were as follows. First, the competency group of the nursing teacher in Specialized vocational high school showed teaching and learning methods & techniques, student guidance, curriculum development and operation, school management support, industry-academy cooperation, self-development and professionalism improvement. The total number was 6. Second, the sub-competency is composed of knowledge and skills related to the curriculum, planning and preparation of instruction, instruction operation, guidance of experiment and practice, instruction evaluation, development and utilization of teaching and learning medium, instruction environment, guidance on student education activities outside the curriculum, career guidance, life guidance, class management, guidance of technology and qualification, development of school curriculum, organization and operation of school curriculum, planning and operation of field practice, work planning, school management support, teacher evaluation and personnel management support, leaning support(practice place and equipment), establishment and operation of industry-academia cooperation, strengthening community cooperation, public relations in the school(department), field research for improvement of educational activities, participation in conferences and training, exploration of new knowledge and technology, The total number was 26. The most important of the relative importance was the curriculum development and operation. The subordinate competencies that have a high priority in each competency group were guidance of experiment and practice, guidance of technology and qualification, planning and operation of field practice, leaning support(practice place and equipment), management and work planning, establishment and operation of industry-academia cooperation, exploration of new knowledge and technology. The competency system derived from this study will be applied to the training and evaluation of nursing teachers in the future and can be used as basic data for related research.

• Journal of Food Hygiene and Safety
• /
• v.34 no.4
• /
• pp.340-347
• /
• 2019

We performed a safety survey on residual pesticides in dried agricultural products. A total of 110 samples of dried agricultural products distributed in Gyeonggi-do were analyzed for 263 pesticides according to multi class pesiticide multiresidue method. Ten types of pesticides were detected in 10 samples. Chlorpyrifos was detected in Ricinus communis leaves, chlorpyrifos, hexaconazole, pyridalyl in Chwinamul (wild aster), diniconazole, isoprothiolane, lufenuron in radish leaves, hexaconazole in Cirsium setidens (Korean thistle), bifenthrin, and chlorothalonil, boscalid, and pyraclostrobin in pepper leaves. The detection rate of pesticides was 9.1%, and among these samples, one was detected over Maximum Residue Limits (MRLs). In the validation study, the values of limit of detection (LOD), limit of quantitation (LOQ), coefficient of determination ($R^2$) and recovery were in the range of 0.002~0.027 mg/kg, 0.006~0.083 mg/kg, 0.9964~1.0000 and 74.8~118.9%, respectively. The Positive List System (PLS) was newly introduced as part of the safety management of residual pesticide in agricultural products in Korea in 2019. With the application of the PLS, if the MRL is not established, 0.01 mg/kg limit is applied uniformly. In spite, these of strengthened residue limits, the MRLs of dried agricultural products are still insufficient. Therefore, this study could be utilized as basic data for the setting of proper MRLs.

• Journal of Intelligence and Information Systems
• /
• v.25 no.3
• /
• pp.239-251
• /
• 2019

Statistical methods such as moving averages, Kalman filtering, exponential smoothing, regression analysis, and ARIMA (autoregressive integrated moving average) have been used for stock market predictions. However, these statistical methods have not produced superior performances. In recent years, machine learning techniques have been widely used in stock market predictions, including artificial neural network, SVM, and genetic algorithm. In particular, a case-based reasoning method, known as k-nearest neighbor is also widely used for stock price prediction. Case based reasoning retrieves several similar cases from previous cases when a new problem occurs, and combines the class labels of similar cases to create a classification for the new problem. However, case based reasoning has some problems. First, case based reasoning has a tendency to search for a fixed number of neighbors in the observation space and always selects the same number of neighbors rather than the best similar neighbors for the target case. So, case based reasoning may have to take into account more cases even when there are fewer cases applicable depending on the subject. Second, case based reasoning may select neighbors that are far away from the target case. Thus, case based reasoning does not guarantee an optimal pseudo-neighborhood for various target cases, and the predictability can be degraded due to a deviation from the desired similar neighbor. This paper examines how the size of learning data affects stock price predictability through k-nearest neighbor and compares the predictability of k-nearest neighbor with the random walk model according to the size of the learning data and the number of neighbors. In this study, Samsung electronics stock prices were predicted by dividing the learning dataset into two types. For the prediction of next day's closing price, we used four variables: opening value, daily high, daily low, and daily close. In the first experiment, data from January 1, 2000 to December 31, 2017 were used for the learning process. In the second experiment, data from January 1, 2015 to December 31, 2017 were used for the learning process. The test data is from January 1, 2018 to August 31, 2018 for both experiments. We compared the performance of k-NN with the random walk model using the two learning dataset. The mean absolute percentage error (MAPE) was 1.3497 for the random walk model and 1.3570 for the k-NN for the first experiment when the learning data was small. However, the mean absolute percentage error (MAPE) for the random walk model was 1.3497 and the k-NN was 1.2928 for the second experiment when the learning data was large. These results show that the prediction power when more learning data are used is higher than when less learning data are used. Also, this paper shows that k-NN generally produces a better predictive power than random walk model for larger learning datasets and does not when the learning dataset is relatively small. Future studies need to consider macroeconomic variables related to stock price forecasting including opening price, low price, high price, and closing price. Also, to produce better results, it is recommended that the k-nearest neighbor needs to find nearest neighbors using the second step filtering method considering fundamental economic variables as well as a sufficient amount of learning data.