• Korean Chemical Engineering Research
• /
• v.57 no.2
• /
• pp.274-282
• /
• 2019

In recent years, big data analysis has been expanded to include automatic control through reinforcement learning as well as prediction through modeling. Research on the utilization of image data is actively carried out in various industrial fields such as chemical, manufacturing, agriculture, and bio-industry. In this paper, we applied NASNet, which is an AutoML reinforced learning algorithm, to DeepU-Net neural network that modified U-Net to improve image semantic segmentation performance. We used BRATS2015 MRI data for performance verification. Simulation results show that DeepU-Net has more performance than the U-Net neural network. In order to improve the image segmentation performance, remove dropouts that are typically applied to neural networks, when the number of kernels and filters obtained through reinforcement learning in DeepU-Net was selected as a hyperparameter of neural network. The results show that the training accuracy is 0.5% and the verification accuracy is 0.3% better than DeepU-Net. The results of this study can be applied to various fields such as MRI brain imaging diagnosis, thermal imaging camera abnormality diagnosis, Nondestructive inspection diagnosis, chemical leakage monitoring, and monitoring forest fire through CCTV.

• KIPS Transactions on Software and Data Engineering
• /
• v.7 no.12
• /
• pp.485-496
• /
• 2018

So far, the deep learning, a field of artificial intelligence, has achieved remarkable results in solving problems from unstructured data. However, it is difficult to comprehensively judge situations like humans, and did not reach the level of intelligence that deduced their relations and predicted the next situation. Recently, deep neural networks show that artificial intelligence can possess powerful relational reasoning that is core intellectual ability of human being. In this paper, to analyze and observe the performance of Relation Networks (RN) among the neural networks for relational reasoning, two types of RN-based deep neural network models were constructed and compared with the baseline model. One is a visual question answering RN model using Sort-of-CLEVR and the other is a text-based question answering RN model using bAbI task. In order to maximize the performance of the RN-based model, various performance improvement experiments such as hyper parameters tuning have been proposed and performed. The effectiveness of the proposed performance improvement methods has been verified by applying to the visual QA RN model and the text-based QA RN model, and the new domain model using the dialogue-based LL dataset. As a result of the various experiments, it is found that the initial learning rate is a key factor in determining the performance of the model in both types of RN models. We have observed that the optimal initial learning rate setting found by the proposed random search method can improve the performance of the model up to 99.8%.

• Korean Journal of Remote Sensing
• /
• v.34 no.6_1
• /
• pp.1055-1066
• /
• 2018

Vegetation indices on the basis of optical characteristics of vegetation can represent various conditions such as canopy biomass and physiological activity. Those have been mostly developed with the large-scaled applications of multi-band optical sensors on-board satellites. However, the sensitivity of vegetation indices for detecting vegetation features will be different depending on the spatial scales. Therefore, in this study, the investigation of photochemical reflectance index (PRI), known as one of useful vegetation indices for detecting photosynthetic ability and vegetation stress, under the three spatial scales was conducted using multi-spectral camera installed in unmanned aerial vehicle (UAV),field spectrometer, and leaf reflectometer. In the leaf scale, diurnal PRI had minimum values at different local-time according to the compass direction of leaf face. It meant that each leaf in some moment had the different degree of light use efficiency (LUE). In early growth stage of crop, $PRI_{leaf}$ was higher than $PRI_{stands}$ and $PRI_{canopy}$ because the leaf scale is completely not governed by the vegetation cover fraction.In the stands and canopy scales, PRI showed a large spatial variability unlike normalized difference vegetation index (NDVI). However, the bias for the relationship between $PRI_{stands}$ and $PRI_{canopy}$ is lower than that in $NDVI_{stands}$ and $NDVI_{canopy}$. Our results will help to understand and utilize PRIs observed at different spatial scales.

• Science of Emotion and Sensibility
• /
• v.22 no.2
• /
• pp.59-69
• /
• 2019

The present study explored consumer's perception and preference on providing information of fashion products by using QR code and suggested the possibility for consumer-to-consumer and consumer-to-company connection. A survey was conducted on males and females in their 20s-a population among whom the rate of smart phone penetration is higher than in any other age group and who tend to exchange information online. The results showed that consumers are dissatisfied with the amount of information, terms of instructions, and ambiguous washing symbols currently provided. Therefore, the study identified the need for better methods of providing information and found that QR code, which is able to deliver high-quality information on fashion products, can be an efficient alternative. Moreover, respondents felt the need for detailed washing instructions, information on handling, and functionality of material on high-involvement fashion products such as outdoor, padding, suit, and underwear worn next to the skin. They also desire styling tips or purchasing information such as SNS OOTD (Outfit Of The Day) utilizing the product, other products that may go well with the one purchased, and similar products on casual wear and coat used on a daily basis. Therefore, QR code used as a link to information web pages or a social network can help consumers to satisfy information needs and to use the products effectively.

• Journal of IKEEE
• /
• v.23 no.2
• /
• pp.438-447
• /
• 2019

As the IoT-based hyper-connected society grows, public-key cryptosystem such as RSA is frequently used for encryption, authentication, and digital signature. Public-key cryptosystem use very large (safe) prime numbers to ensure security against malicious attacks. Even though the performance of the device has greatly improved, the generation of a large (safe)prime is time-consuming or memory-intensive. In this paper, we propose ET-MR and ET-MR-MR using Euler sieve so it runs faster while using less memory. We present a running time prediction model by probabilistic analysis and compare time and memory of our method with conventional methods. Experimental results show that the difference between the expected running time and the measured running time is less than 4%. In addition, the fastest running time of ET-MR is 36% faster than that of TD-MR, 8.5% faster than that of DT-MR and the fastest running time of ET-MR-MR is 65.3% faster than that of TD-MR-MR and similar to that of DT-MR-MR. When k=12,381, the memory usage of ET-MR is 2.7 times more than that of DT-MR but 98.5% less than that of TD-MR and when k=65,536, the memory usage of ET-MR-MR is 98.48% less than that of TD-MR-MR and 92.8% less than that of DT-MR-MR.

• Biomolecules & Therapeutics
• /
• v.27 no.5
• /
• pp.474-483
• /
• 2019

Vascular endothelial growth factor (VEGF) plays a pivotal role in pathologic ocular neovascularization and vascular leakage via activation of VEGF receptor 2 (VEGFR2). This study was undertaken to evaluate the therapeutic mechanisms and effects of the tetrapeptide Arg-Leu-Tyr-Glu (RLYE), a VEGFR2 inhibitor, in the development of vascular permeability and choroidal neovascularization (CNV). In cultured human retinal microvascular endothelial cells (HRMECs), treatment with RLYE blocked VEGF-A-induced phosphorylation of VEGFR2, Akt, ERK, and endothelial nitric oxide synthase (eNOS), leading to suppression of VEGF-A-mediated hyper-production of NO. Treatment with RLYE also inhibited VEGF-A-stimulated angiogenic processes (migration, proliferation, and tube formation) and the hyperpermeability of HRMECs, in addition to attenuating VEGF-A-induced angiogenesis and vascular permeability in mice. The anti-vascular permeability activity of RLYE was correlated with enhanced stability and positioning of the junction proteins VE-cadherin, ${\beta}$-catenin, claudin-5, and ZO-1, critical components of the cortical actin ring structure and retinal endothelial barrier, at the boundary between HRMECs stimulated with VEGF-A. Furthermore, intravitreally injected RLYE bound to retinal microvascular endothelium and inhibited laser-induced CNV in mice. These findings suggest that RLYE has potential as a therapeutic drug for the treatment of CNV by preventing VEGFR2-mediated vascular leakage and angiogenesis.

• Journal of Navigation and Port Research
• /
• v.43 no.1
• /
• pp.72-78
• /
• 2019

The second-hand ship market provides immediate access to the freight market for shipping investors. When introducing second-hand vessels, the precise estimate of the price is crucial to the decision-making process because it directly affects the burden of capital cost to investors in the future. Previous studies on the second-hand market have mainly focused on the market efficiency. The number of papers on the estimation of second-hand vessel values is very limited. This study proposes an artificial neural network model that has not been attempted in previous studies. Six factors, freight, new-building price, orderbook, scrap price, age and vessel size, that affect the second-hand ship price were identified through literature review. The employed data is 366 real trading records of Panamax second-hand vessels reported to Clarkson between January 2016 and December 2018. Statistical filtering was carried out through correlation analysis and stepwise regression analysis, and three parameters, which are freight, age and size, were selected. Ten-fold cross validation was used to estimate the hyper-parameters of the artificial neural network model. The result of this study confirmed that the performance of the artificial neural network model is better than that of simple stepwise regression analysis. The application of the statistical verification process and artificial neural network model differentiates this paper from others. In addition, it is expected that a scientific model that satisfies both statistical rationality and accuracy of the results will make a contribution to real-life practices.

• The Bulletin of The Korean Astronomical Society
• /
• v.43 no.2
• /
• pp.36.2-36.2
• /
• 2018

Cosmic reionization era in the early universe was playing a leading part on making the present universe we know. However, we have not been able to reveal the main contributor to the cosmic reionization to date. Faint quasars have been mentioned as the alternative due to the uncertainty of the faint end slope of the quasars luminosity function. With the availability of the deep (~25mag) images from Subaru Hyper Suprime-Cam (HSC) Strategic Program survey, we have tried to find more quasar with low luminosity in the ELAIS-N1 field. Faint quasar candidates were selected from several multi-band color cut criteria based on the track of the simulated quasar at z ~ 5. The Infrared Medium-deep Survey (IMS) and The UKIRT Infrared Deep Sky Survey (UKIDSS) - Deep Extragalactic Survey (DXS) provide J band information which is used to cover the relatively long wavelength range of quasar spectra. To search the reliable candidates with possible Lyman break, medium-band observation was performed by the SED camera for QUasars in EArly uNiverse(SQUEAN) in the McDonald observatory and Seoul National University 4k Camera(SNUCAM) in the Maidanak observatory. Photometric redshifts of the observed candidates were estimated from chi-square minimization. Also, we predicted the importance of the faint quasar to the cosmic reionization from the expected number density of the faint quasar.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.20 no.5
• /
• pp.36-46
• /
• 2019

The number of hosts connected to the Internet has increased dramatically, introducing the Domain Name System(DNS) in 1984. DNS is now an important key point for all users of the Internet by allowing them to use a convenient character address without memorizing a series of numbers of complex IP address. However, relative to the importance of DNS, there still exist many problems such as the authorization allocation issue, the disputes over public registration, security vulnerability such as DNS cache poisoning, DNS spoofing, man-in-the-middle attack, DNS amplification attack, and the need for many domain names in the age of hyper-connected networks. In this paper, to effectively improve these problems of existing DNS, we proposed a method of implementing DNS using distributed ledger technology, blockchain, and implemented using a Ethereum-based platform. In addition, the qualitative analysis performance comparative evaluation of the existing domain name registration and domain name server was conducted, and conducted security assessments on the proposed system to improve security problem of existing DNS. In conclusion, it was shown that DNS services could be provided high security and high efficiently using blockchain.

• Korean Journal of Clinical Laboratory Science
• /
• v.51 no.2
• /
• pp.221-234
• /
• 2019

Cancer is genetically, metabolically and infectiously induced life threatening disorder showing aggressive growing pattern with invasive tendency. In order to prevent this global menace from jeopardizing human life, enormous studies on carcinogenesis and treatment for chemotherapy resistance have been intensively researched. Hinokitiol (${\beta}$-thujaplicin) extracted from heart wood of cupressaceous is a well-known bioactive compound demonstrating anti-inflammation, anti-bacteria and anti-cancer effects on several cancer types via apoptosis and autophagy. This study proposed that hinokitiol activates transcription factor EB (TFEB) nuclear translocation for autophagy and lysosomal biogenesis regardless of nutrient condition in cancer cells. Mitophagy and ${\beta}$-catenin translocation into the nucleus under treatment of hinokitiol on non-small cell lung cancer (NSCLC) cells and HeLa cells were investigated. Hinokitiol exerted cytotoxicity on HeLa and HCC827 cells; moreover, artificially induced autophagy by overexpression of TFEB granted imperfect sustainability onto HeLa cells. Taken together, hinokitiol is the prominent autophagy inducer and activator of TFEB nuclear translocation. Alternative cancer therapy via autophagy is pros and cons since the autophagy in cancer cells is related to prevention and survival mechanism depending on nutrition. To avoid paradox of autophagy in cancer therapy, fine-tuned regulation and application of hinokitiol in due course for successful suppressing cancer cells are recommended.