• Communications for Statistical Applications and Methods
• /
• v.27 no.3
• /
• pp.313-326
• /
• 2020

Deep Learning is the most important key to the development of Artificial Intelligence (AI). There are several distinguishable architectures of neural networks such as MLP, CNN, and RNN. Among them, we try to understand one of the main architectures called Recurrent Neural Network (RNN) that differs from other networks in handling sequential data, including time series and texts. As one of the main tasks recently in Natural Language Processing (NLP), we consider Neural Machine Translation (NMT) using RNNs. We also summarize fundamental structures of the recurrent networks, and some topics of representing natural words to reasonable numeric vectors. We organize topics to understand estimation procedures from representing input source sequences to predict target translated sequences. In addition, we apply multiple translation models with Gated Recurrent Unites (GRUs) in Keras on English-Korean sentences that contain about 26,000 pairwise sequences in total from two different corpora, colloquialism and news. We verified some crucial factors that influence the quality of training. We found that loss decreases with more recurrent dimensions and using bidirectional RNN in the encoder when dealing with short sequences. We also computed BLEU scores which are the main measures of the translation performance, and compared them with the score from Google Translate using the same test sentences. We sum up some difficulties when training a proper translation model as well as dealing with Korean language. The use of Keras in Python for overall tasks from processing raw texts to evaluating the translation model also allows us to include some useful functions and vocabulary libraries as well.

• Journal of Korean Society of Water and Wastewater
• /
• v.19 no.6
• /
• pp.738-748
• /
• 2005

Considering pollution source is transferred by discharge, it is very important to analyze the correlation between discharge and water quality. And temperature also influent to the water quality. In this paper, it is used water quality data that was measured DO (Dissolved Oxygen), TOC (Total Organic Carbon), TN (Total Nitrogen), TP (Total Phosphorus) at Dalchun real time monitoring stations in Namhan river. These characteristics were analyzed with the water quality of rainy and nonrainy periods. Input data of the water quality forecasting models that they were constructed by neural network and neuro-fuzzy was chosen as the reasonable data, and water quality forecasting models were applied. LMNN (Levenberg-Marquardt Neural Network), MDNN (MoDular Neural Network), and ANFIS (Adaptive Neuro-Fuzzy Inference System) models have achieved the highest overall accuracy of TOC data. LMNN and MDNN model which are applied for DO, TN, TP forecasting shows better results than ANFIS. MDNN model shows the lowest estimation error when using daily time, which is qualitative data trained with quantitative data. If some data has periodical properties, it seems effective using qualitative data to forecast.

• International journal of advanced smart convergence
• /
• v.9 no.2
• /
• pp.20-27
• /
• 2020

Sound event detection in real-world environments suffers from the interference of non-stationary and time-varying noise. This paper presents an adaptive noise reduction method for sound event detection based on non-negative matrix factorization (NMF). In this paper, we proposed a deep learning model that integrates Convolution Neural Network (CNN) with Non-Negative Matrix Factorization (NMF). To improve the separation quality of the NMF, it includes noise update technique that learns and adapts the characteristics of the current noise in real time. The noise update technique analyzes the sparsity and activity of the noise bias at the present time and decides the update training based on the noise candidate group obtained every frame in the previous noise reduction stage. Noise bias ranks selected as candidates for update training are updated in real time with discrimination NMF training. This NMF was applied to CNN and Hidden Markov Model(HMM) to achieve improvement for performance of sound event detection. Since CNN has a more obvious performance improvement effect, it can be widely used in sound source based CNN algorithm.

• Biotechnology and Bioprocess Engineering:BBE
• /
• v.5 no.4
• /
• pp.263-274
• /
• 2000

Carotenoids are widely distributed natural pigments which are in an increasing demand by the market, due to their applicatins in the human food, animal feed, cosmetics, and pharmaceutical industries. Although more than 600 carotenoids have been identified in nature, only a few are industrially important (${\beta}$-carotene, astaxanthin, lutein or lycopene). To date chemical processes manufacture most of the carotenoid production, but the interest for carotenoids of biological origin is growing since theire is an increased public concern over the safety of artificial food colorants. Although much interest and effort has been devoted to the use of biological sources for industrially important carotenoids, only the production of biological ${\beta}$-carotene and astaxanthin has been reported. Among fungi, several Mucorales strains, particularly Blakeslea trispora, have been used to develop fermentation processes for the production of ${\beta}$-carotene on almost competitive cost-price levels. Similarly, the basidiomycetous yeast Xanthophyllomyces dendrorhous (the perfect state of Phaffia rhodozyma), has been proposed as a promising source of astaxanthin. This paper focuses on recent findings on the fungal pathways for carotenoid production, especially the structure and function of the genes involved in the biosynthesis of carotenoids in the Mucorales. An outlook of the possibilities of an increased industrial production of carotenoids, based on metabolic engineering of fungi for carotenoid content and composition, is also discussed.

• The Research Journal of the Costume Culture
• /
• v.15 no.1 s.66
• /
• pp.68-80
• /
• 2007

Surrealism's method of expression and inspiration has been a continuous influence on fashion. What is important is how to identify and research a source of inspiration, and how to use this inspiration to guide the designs that can be worn on a real human body. Therefore, the purpose of this study was to show how to develop ideas and designs from Surrealist objects and how to contribute to the development of designs to the diversity in visual and functional effects along with the changes in consumer's interest. Documentary studies about Surrealism and Surrealist-inflected fashion applying the works of Surrealist and fashion images as a tool of historical research were proceeded. And the ideas for theme, the motifs for print patterns, and the concept of designs were borrowed from Surrealist objects touching on the imagery between the real and reflected, optical and artificial, and the designs were developed with several different visual and functional effects changing color, pattern, size, and length with the function of zippers in each item. It also gives the person wearing it amusement because it includes the concept of play and D.I.Y. With effective planning, it is possible to attract consumer's interest in the market.

• Journal of computational fluids engineering
• /
• v.18 no.1
• /
• pp.43-48
• /
• 2013

The feasibility of a unique greenhouse, named as Gradient Biome, is now being examined extensively in the University of Tokyo. It is a large chamber (length:200m, width:50m, height:40m) in which the weather, such as temperature and humidity, of the tropical zone through to that of the frigid zone on the earth is reproduced with continuous gradient. In the Gradient Biome, ecosystems (mainly plants) corresponding to each weather are introduced and the possible responses of this ecosystems to the expected global warming are to be observed. Since one of the expected responses is the shift of the ecosystem(s) toward the region of suitable environment, there should be no artificial obstacles, which can prevent the shift, inside the Biome. This requirement is not so easy to be satisfied since the environment tends to be homogeneous. This paper presents the results of the numerical studies conducted to find the ways of how the temperature and humidity in the Gradient Biome could be reproduced. One of the contributed solvers of OpenFOAM, which is an open source physics simulation code, was modified and used for the numerical simulations.

• Korean Journal of Fisheries and Aquatic Sciences
• /
• v.36 no.5
• /
• pp.469-473
• /
• 2003

Transplantation of Sargassum horneri to natural substrate using the rope seeding method was undertaken. Seeding of S. horneri was grown on net in an indoor tank up to 0.2-0.3 cm in frond length for 2 months, and then transplanted to an architecture tile $(10\times10\;cm)$ with underwater glue and an iron pipe. After two months, number of S. horneri on the tile decreased because of grazing by herbivores. However, S. horneri on the iron pipe grew up to 0.5-2.0 cm in frond length. After 7 months, they rapidly grew up to a frond length of 11.0-203.0 cm with a mean frond length of 122.6 cm. S. horneri grew up to a maximum frond length of 313.0 cm, and mean frond length of 228.0 cm, after 10 months. The artificial S. hornevi beds can be used to clean the seawater and also be utilized as a source of marine biomass as well as for supplying habitat, shelter and spawning beds for fish and invertebrates.

• Korean Journal of Fisheries and Aquatic Sciences
• /
• v.43 no.6
• /
• pp.687-692
• /
• 2010

The mass cultivation of Ecklonia cava Kjellman was studied as a potential biomass source for the extract industry in Korea. Experiments were conducted to investigate the optimal conditions for artificial seed production and mass cultivation of this species. Maximum growth and young thalli development in the nursery culture area occurred at 2 m depth, whereas maximum growth of thalli in the main culture area occurred at 1 m depth. Production of E. cava was between 2.6 and 3.6 kg wet wt. $m^{-1}$ after depth control and removal of fouling organism, etc. The relationship between optimal water depth for culture and underwater irradiance during the E. cava cultivation was calculated as: y = -0.718x + 8.042 ($r^2$=0.976). The growth rates achieved in this trial indicate that E. cava cultures could produce and supply sufficient biomass.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.9 no.11
• /
• pp.4483-4501
• /
• 2015

In this paper, we investigate a security transmission scheme at the physical layer for cooperative wireless relay networks in the presence of a passive eavesdropper. While the security scheme has been previously investigated with perfect channel state information(CSI) in the presence of a passive eavesdropper, this paper focuses on researching the robust cooperative relay beamforming mechanism for wireless relay networks which makes use of artificial noise (AN) to confuse the eavesdropper and increase its uncertainty about the source message. The transmit power used for AN is maximized to degrade the signal-to-interference-plus-noise-ratio (SINR) level at the eavesdropper, while satisfying the individual power constraint of each relay node and worst-case SINR constraint at the desired receiver under a bounded spherical region for the norm of the CSI error vector from the relays to the destination. Cooperative beamforming weight vector in the security scheme can be obtained by using S-Procedure and rank relaxation techniques. The benefit of the proposed scheme is showed in simulation results.

• Microbiology and Biotechnology Letters
• /
• v.22 no.4
• /
• pp.415-422
• /
• 1994

The biodegradation of aromatic compounds by a mixed culture GE1 was investigated in an artificial wastewater containing 250 mg/l of benzene, toluene, and phenol in semicontinuous culture. In the control group (no strains) with an aeration rate of 75 ml/l/min, 37% of phenol and 83% of benzene were volatilized during early 24 hrs and toluene was disappeared from the medium within 12 hrs. The biodegradation of benzene and toluene was effective in SB (strains + biofilm) treatment, while phenol was degraded more quickly in SG (strains + glucose) treatment including glucose as an additional carbon source. aromatic compounds added at a concentration of 250 mg/l were completely removed by SG treatment after 16 hrs or 32 hrs, respectively. The removal rate of COD was high as much as 80 mg/l/h in SG treatment during early period, but COD revealed a stable value of 116~140 mg/l after 12 hrs caused by increased biomass. Therefore, it is concluded that the mixed GE1 could be used for the wastewater treatment including aromatic compounds such as benzene, toluene, and phenol.