• Journal of the Korean Data and Information Science Society
• /
• v.28 no.2
• /
• pp.287-295
• /
• 2017

Since Google's AlphaGo defeated a world champion of Go players in 2016, there have been many interests in the deep learning. In the financial sector, a Robo-Advisor using deep learning gains a significant attention, which builds and manages portfolios of financial instruments for investors.In this paper, we have proposed the a deep learning algorithm geared toward identification and forecast of the KOSPI index direction,and we also have compared the accuracy of the prediction.In an application of forecasting the financial market index direction, we have shown that the Robo-Advisor using deep learning has a significant effect on finance industry. The Robo-Advisor collects a massive data such as earnings statements, news reports and regulatory filings, analyzes those and recommends investors how to view market trends and identify the best time to purchase financial assets. On the other hand, the Robo-Advisor allows businesses to learn more about their customers, develop better marketing strategies, increase sales and decrease costs.

• The Bulletin of The Korean Astronomical Society
• /
• v.36 no.2
• /
• pp.121-121
• /
• 2011

Thanks to KMTNet's wide field of view, it is time to implement imaging survey of extensive area of clusters of galaxies in the southern sky with modern instrument. As part of potential long-term survey of nearby (D < 50 Mpc) well-known clusters of galaxies, we propose a wide-field and deep survey of Fornax cluster as a first step of the project. By imaging the 400 square deg region (100 fields) enclosed within the five times virial radius of the Fornax cluster, in three SDSSfilters(g', r', i'), we can provide an unprecedented view of structure of Fornax cluster using sample from giant to dwarf galaxies. We will secure galaxies with brightness comparable to the limiting magnitude (r'=23.1 AB mag) of SDSS. Furthermore, we also request extremely deep (limiting surface brightness of ~ 28 mag $arcsec^{-2}$forr'band) survey for the central region (16 square degree, i.e., four fields) of Fornax cluster. This will allow us to detect the diffuse intracluster light (ICL) that permeates clusters as a valuable tool for studying the hierarchical nature of cluster assembly. In order to complete whole survey, about 285 hr observing time (without overhead) is required. By combining data available at other wavelengths, it will offer unique constraints on the formation of large-scale structure and also provide important clues for theories of galaxy formation and evolution. Our proposed survey will be implemented in the close collaboration with researchers in various countries (Germany, Australia, UK, USA) and ongoing project (e.g., SkyMapper).

• Journal of the Architectural Institute of Korea Planning & Design
• /
• v.35 no.12
• /
• pp.43-53
• /
• 2019

The construction industry has made great strides in the past decades by utilizing computer programs including CAD. However, compared to other manufacturing sectors, labor productivity is low due to the high proportion of workers' knowledge-based task in addition to simple repetitive task. Therefore, the knowledge-based task efficiency of workers should be improved by recognizing the visual information of computers. A computer needs a lot of training data, such as the ImageNet project, to recognize visual information. This study, aim at proposing building facade datasets that is efficiently constructed by quickly collecting building facade data through portal site road view and automatically labeling using deep learning as part of construction of image dataset for visual recognition construction by the computer. As a method proposed in this study, we constructed a dataset for a part of Dongseong-ro, Daegu Metropolitan City and analyzed the utility and reliability of the dataset. Through this, it was confirmed that the computer could extract the significant facade information of the portal site road view by recognizing the visual information of the building facade image. Additionally, In contribution to verifying the feasibility of building construction image datasets. this study suggests the possibility of securing quantitative and qualitative facade design knowledge by extracting the facade design knowledge from any facade all over the world.

• Proceedings of the Korean Society of Crop Science Conference
• /
• 2022.10a
• /
• pp.80-80
• /
• 2022

Estimation of heading date of paddy rice is laborious and time consuming. Therefore, automatic estimation of heading date of paddy rice is highly essential. In this experiment, deep learning classification models were used to classify two difference categories of rice (vegetative and reproductive stage) based on the panicle initiation of paddy field. Specifically, the dataset includes 444 slanted view images belonging to two categories and was then expanded to include 1,497 images via IMGAUG data augmentation technique. We adopt two transfer learning strategies: (First, used transferring model weights already trained on ImageNet to six classification network models: VGGNet, ResNet, DenseNet, InceptionV3, Xception and MobileNet, Second, fine-tuned some layers of the network according to our dataset). After training the CNN model, we used several evaluation metrics commonly used for classification tasks, including Accuracy, Precision, Recall, and F1-score. In addition, GradCAM was used to generate visual explanations for each image patch. Experimental results showed that the InceptionV3 is the best performing model in terms of the accuracy, average recall, precision, and F1-score. The fine-tuned InceptionV3 model achieved an overall classification accuracy of 0.95 with a high F1-score of 0.95. Our CNN model also represented the change of rice heading date under different date of transplanting. This study demonstrated that image based deep learning model can reliably be used as an automatic monitoring system to detect the heading date of rice crops using CCTV camera.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.6
• /
• pp.1833-1848
• /
• 2022

Named entity recognition (NER) is an important basic task in the field of Natural Language Processing (NLP). Recently deep learning approaches by extracting word segmentation or character features have been proved to be effective for Chinese Named Entity Recognition (CNER). However, since this method of extracting features only focuses on extracting some of the features, it lacks textual information mining from multiple perspectives and dimensions, resulting in the model not being able to fully capture semantic features. To tackle this problem, we propose a novel Multi-view Semantic Feature Fusion Model (MSFM). The proposed model mainly consists of two core components, that is, Multi-view Semantic Feature Fusion Embedding Module (MFEM) and Multi-head Self-Attention Mechanism Module (MSAM). Specifically, the MFEM extracts character features, word boundary features, radical features, and pinyin features of Chinese characters. The acquired font shape, font sound, and font meaning features are fused to enhance the semantic information of Chinese characters with different granularities. Moreover, the MSAM is used to capture the dependencies between characters in a multi-dimensional subspace to better understand the semantic features of the context. Extensive experimental results on four benchmark datasets show that our method improves the overall performance of the CNER model.

• Childhood Kidney Diseases
• /
• v.11 no.2
• /
• pp.152-160
• /
• 2007

Twenty-six renal allograft biopsies which showed acute rejection and had renal capsule and medulla in the same specimen were selected in order to compare the severity of acute rejection between superficial cortex, deep cortex and medulla. Disregarding the mid cortical region, the superficial cortex was considered as being one-third of the distance from the renal capsule to the medulla and the deep cortex as being that one-third of the cortex which was adjacent to the medulla. Using semiquantitative histologic analysis the following parameters were compared in superficial cortex, deep cortex, and medulla: interstitial inflammation, edema, tubulitis, and acute tubulointerstitial rejection grade. Also, the presence of lymphocyte activation and polymorphonuclear leukocytes was evaluated. Significantly greater histologic changes of acute rejection were found in the deep cortex vs. supeficial cortex for the following parameters: interstitial inflammation(P=0.013), edema (P=0.023) and tubulointerstitial rejection grade(P=0.016). These findings support the view that biopsies in which deep cortex is not included may result in underestimation of the severity of renal allograft rejection.

• Cross-Cultural Studies
• /
• v.42
• /
• pp.47-68
• /
• 2016

This paper closely examines Popol Vuh, regarded as the Bible of the Mayans, from the perspective of the Deep Ecology. Deep ecology is a concept born out of the criticisms on the existing environmental movements as superficial, and encourages the moral and ethical change of the man's attitude towards the nature, inevitably becoming "metaphysical" in character. As such the Deep ecology advocates the break away from the anthropocentricism, the dichotomous thinking and the rationalism of the modern times. Popol Vuh is a text that contains such concepts and it requires an analysis from the Deep ecological perspective beyond the existing framework of the study that simply focuses on mythological elements.

• Smart Structures and Systems
• /
• v.31 no.5
• /
• pp.469-483
• /
• 2023

The dynamic characteristics of wind turbine blades are usually monitored by contact sensors with the disadvantages of high cost, difficult installation, easy damage to the structure, and difficult signal transmission. In view of the above problems, based on computer vision technology and the improved YOLOv5 (You Only Look Once v5) deep learning model, a non-contact dynamic characteristic monitoring method for wind turbine blade is proposed. First, the original YOLOv5l model of the CSP (Cross Stage Partial) structure is improved by introducing the CSP2_2 structure, which reduce the number of residual components to better the network training speed. On this basis, combined with the Deep sort algorithm, the accuracy of structural displacement monitoring is mended. Secondly, for the disadvantage that the deep learning sample dataset is difficult to collect, the blender software is used to model the wind turbine structure with conditions, illuminations and other practical engineering similar environments changed. In addition, incorporated with the image expansion technology, a modeling-based dataset augmentation method is proposed. Finally, the feasibility of the proposed algorithm is verified by experiments followed by the analytical procedure about the influence of YOLOv5 models, lighting conditions and angles on the recognition results. The results show that the improved YOLOv5 deep learning model not only perform well compared with many other YOLOv5 models, but also has high accuracy in vibration monitoring in different environments. The method can accurately identify the dynamic characteristics of wind turbine blades, and therefore can provide a reference for evaluating the condition of wind turbine blades.

• Journal of Nuclear Fuel Cycle and Waste Technology(JNFCWT)
• /
• v.16 no.4
• /
• pp.491-505
• /
• 2018

As an alternative to deep geological disposal technology, which is considered as a reference concept, the domestic applicability of deep borehole disposal technology for high level radioactive waste, including spent fuel, has been preliminarily evaluated. Usually, the environment of deep borehole disposal, at a depth of 3 to 5 km, has more stable geological and geo-hydrological conditions. For this purpose, the characteristics of rock distribution in the domestic area were analyzed and drilling and investigation technologies for deep boreholes with large diameter were evaluated. Based on the results of these analyses, design criteria and requirements for the deep borehole disposal system were reviewed, and preliminary reference concept for a deep borehole disposal system, including disposal container and sealing system meeting the criteria and requirements, was developed. Subsequently, various performance assessments, including thermal stability analysis of the system and simulation of the disposal process, were performed in a 3D graphic disposal environment. With these analysis results, the preliminary evaluation of the domestic applicability of the deep borehole disposal system was performed from various points of view. In summary, due to disposal depth and simplicity, the deep borehole disposal system should bring many safety and economic benefits. However, to reduce uncertainty and to obtain the assent of the regulatory authority, an in-situ demonstration of this technology should be carried out. The current results can be used as input to establish a national high-level radioactive waste management policy. In addition, they may be provided as basic information necessary for stakeholders interested in deep borehole disposal technology.

• Proceedings of the Korea Association of Crystal Growth Conference
• /
• 1997.06a
• /
• pp.253-259
• /
• 1997

The iron doped GaAs single crystals were grown by liquid phase epitaxial method and its some physical properties were evaluated with a view to investigate the crystal quality and emission property. The isomer shift of 0.303mm/sec is calculated from low-temperature M ssbauer spectroscopy and we know that charge state of iron ion is 3+ in GaAs crystal. In low temperature photoluminescence, the deep emission bands with wide-line width have been observed at 0.99eV and 1.15eV in addition to sharp excitonic peaks. We attributed that these deep emissions are originated from substitutional Fe-acceptor which has charge state of 3+ and 2+, respectively.