• The Journal of Asian Finance, Economics and Business
• /
• 제7권10호
• /
• pp.43-50
• /
• 2020

Disclosure level in interim financial reporting is important for information users to make business decisions. This has received much attention from the information users. The article is aimed at determining the factors of financial ratios, which impact on the disclosure level in interim financial reporting. The authors use the ordinary least squares to test. The sample consists of 418 VN100 over a 6-year period from 2014 to 2019. The results show that there are four factors that positively impact on the disclosure level in interim financial reporting: Enterprise size (SIZE); Liquidity (LIQI); Sales growth (GROW) and Profitability (ROE). The article proposes some policy recommendations to contribute to improving disclosure level in interim financial reporting. Accordingly, State Securities Commission of Vietnam should strengthen the regular inspection of VN100's disclosure level in interim financial reporting and also should enforce strict sanctions or may consider delisting in cases of listed enterprises with incomplete disclosure. The managers of VN100 need to raise the sense of responsibility of information providers to ensure adequate information in interim financial reporting. Investors should also pay attention to the financial ratios of VN100 such as firm size, return-on-equity, liquidity, and sales growth to get useful information and ensure sound business decisions.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제16권3호
• /
• pp.877-893
• /
• 2022

With the development of deep learning, face inpainting has been significantly enhanced in the past few years. Although image inpainting framework integrated with generative adversarial network or attention mechanism enhanced the semantic understanding among facial components, the issues of reconstruction on corrupted regions are still worthy to explore, such as blurred edge structure, excessive smoothness, unreasonable semantic understanding and visual artifacts, etc. To address these issues, we propose a Learnable Structure Knowledge of Fusion Network (LSK-FNet), which learns a prior knowledge by edge generation network for image inpainting. The architecture involves two steps: Firstly, structure information obtained by edge generation network is used as the prior knowledge for face inpainting network. Secondly, both the generated prior knowledge and the incomplete image are fed into the face inpainting network together to get the fusion information. To improve the accuracy of inpainting, both of gated convolution and region normalization are applied in our proposed model. We evaluate our LSK-FNet qualitatively and quantitatively on the CelebA-HQ dataset. The experimental results demonstrate that the edge structure and details of facial images can be improved by using LSK-FNet. Our model surpasses the compared models on L1, PSNR and SSIM metrics. When the masked region is less than 20%, L1 loss reduce by more than 4.3%.

• 정보관리학회지
• /
• 제22권4호통권58호
• /
• pp.5-22
• /
• 2005

광고산업과 미디어의 발전에 따라 광고물의 증가뿐만 아니라 광고와 관련한 많은 연구가 이루어지고 있으며, 그 결과 광고 관련 자료가 꾸준히 증가하고 있는 추세이다. 그런데 KDC 광고분야의 분류체계를 살펴보면 광고관련 분류항목이 너무나 미비하며, 더욱이 주류 배치에 있어 관련 학문과의 연관성을 고려하지 않아 사서나 이용자 모두에게 혼란을 초래하고 있다. 이러한 점을 감안하여 본 연구에서는 광고의 유형 및 학문적 특성에 대해 고찰하고, KDC를 비롯한 NDC, DBC, LCC와 같은 문헌분류법 및 광고 전문도서관, 그리고 인터넷 포털사이트와 인터넷서점 등의 분류체계를 분석함으로써 KDC 광고분야의 분류체계에 대한 문제점을 알아보고 이에 대한 개선방안을 제시하였다.

• 한국컴퓨터산업학회논문지
• /
• 제3권5호
• /
• pp.575-584
• /
• 2002

온라인 필기체 문자 인식은 필기의 순서와 획의 위치를 알 수 있어 신경망을 이용한 자소의 효과적인 분할로 큰 성과를 이루었다. 그러나 오프라인 필기체 문자 인식은 동적인 정보와 시간적인 정보를 가지고 있지 않고, 다양한 필기와 자소의 겹침이 심하며 획 사이의 잡영을 많이 가지고 있어 불완전한 전처리를 수행하여야 하는 어려움을 가지고 있다. 따라서 오프라인 필기체 문자 인식은 다양한 방법의 연구가 필요하다. 본 논문에서는 Watershed 알고리즘을 오프라인 필기체 한글 문자 인식 전처리에 적용하였다. 여기서 Watershed 알고리즘의 수행 시간과 결과 영상의 품질을 고려해 Watershed 알고리즘 4단계에서 효과적인 적용방법을 제시하였다. 효과적으로 구성된 Watershed 알고리즘을 전처리에 적용함으로써 영상 향상과 이진화에 좋은 결과를 얻었다. 실험에서는 기존의 방법과 본 논문 방법을 수행 시간과 품질로써 평가했다. 실험 결과 기존의 방법은 평균 2.08초, 본 논문 방법은 평균 0.86초의 수행 시간이 걸렸다. 결과 영상의 품질은 본 논문 방법이 기존의 방법에 비하여 문자의 획 사이의 잡영을 효과적으로 처리하였다.

• 한국학교수학회논문집
• /
• 제1권1호
• /
• pp.97-109
• /
• 1998

The activities of teaching and learning are to try to reach the lesson object most closely in many ways. Considering that the lesson objects are to get the principle or law of a concept, to acquire the mathematical function, to master it through repeated exercises and to solve mathematical problems, we need many ways to reach such objects. Among the many ways, we can first think of one: the students will learn with curiosity and according to their own ability or advancing level in learning when teachers study and prepare necessary contents enough in advance by using computers, showing the right program to learners' needs. For example, defining definite integral by measuration by parts will help understand measuration by parts well and know the meaning of definite integral correctly, In teaching and learning by the use of this program, the educational effects are expected as follows. 1. It is thought that this program will stimulate the desire for and interest in learning because it used animation and acoustic effect. And voluntary and positive thinking activity will be shown. 2. It is expected that the conviction of formulas will be got and the concept of definite integral will be remembered firmly by showing how to measure the width of circle with the use of measuration by parts in various other ways instead of the ways used at present. 3. It is expected that students will feel the pleasure of mathematics in life when they recognize mathematical facts scattered really in our life rather than mathematical difficulties. 4. It is expected that the repeated review of programs already designed will remove the fear of incomplete parts and help review again. 5. It is certain that positive attitude in life will be formed as teacher-centered class is changed into learner-centered class and unwilling study is changed into self-oriented study. However, I think this program is insufficient for humanbeing-centered education given directly in contact with students on the ground of the variety in mathematical education and applications in many ways. And mechanically inhuman computers leave some solutions to be desired

• 디자인학연구
• /
• 제19권3호
• /
• pp.207-218
• /
• 2006

21st century is called digital and information era. The visual communication becomes more important than any other centuries. Visual communication design has been represented in varied images from ancient society up until the present time. Especially graphic design, which has experienced artistic upheaval from modernism and postmodernism, has been enabled to deliver messages in a more developed and sensuous appearances. Deconstructionist design wasn't well received by modernism, an advocate of practicality and effectiveness; however, as the beginning of subsequent postmodernism, it started to earn a warm welcome from a number of designers. Postmodernism made it possible for deconstruction to be a new paradigm appreciating neutral respect for varied expressions and designers' idiosyncrasy. Corresponding to this new trend, Edward Fella published his artwork 'Letters on America', which introduced a variety of vernacular design works in America. It was the result of years of research, and included materials which were not something completely new to field of communication design but rather symbolic, deconstructive images available everywhere around us. Deconstruction broke away from the previous modernism's unity and standards, and sought to achieve a new tensional deliverance. As a consequence, images were created in destructive and incomplete forms. Graphic design also complied with this new code and its advent fascinated the artists in the fields of advertising, editorial and poster designs. This paper contends that contemporary graphic design still maintains a deconstructive representation, examines its locality in design works from an aesthetic viewpoint, and researches theoretical approaches to deconstructionist and case studies conducted domestically. There has been incessant critical contradiction in terms of the heritage of deconstruction. I would like to present you with a moment in which you can struggle to decide whether it is a result of an effort to get away from modernism or an offspring of postmodernism.

• 전자공학회논문지C
• /
• 제36C권6호
• /
• pp.44-51
• /
• 1999

HMM-Net은 은닉 마르코프 모델(HMM)의 계산과정을 신경망 구조로 구현한 것으로, HMM이 갖고 있는 시계열 모델링 능력과 신경망이 갖고 있는 우수한 변별력을 결합한 것이다. HMM-Net 분류기를 학습하는 방법으로는 HMM의 학습에 이용되는 最尤法(ML)과 신경망 학습의 最小二乘誤差法(MMSE)를 적용할 수 있다. 이들 중 ML이 MMSE보다 안정된 학습을 보장하는 반면 초기 학습조건을 적절하게 설정하였을 경우에는 MMSE가 ML보다 우수하다고 알려져 있다[3]. 따라서 이 논문에서는 먼저 ML을 이용하여 초기학습을 수행한 다음 보다 학습성능이 우수한 MMSE로 바꾸어 최적 또는 준최적으로 학습하는 하이브리드 학습법(ML/MMSE)을 제안한다. 실험용 시계열 패턴으로 /0/부터 /9/까지의 고립 숫자음을 이용하여 실험한 결과, 제안한 방법이 학습특성 및 인식률면에서 ML이나 MMSE만을 이용하는 기존의 방법보다 우수하였음을 확인하였다.

• 한국항해항만학회:학술대회논문집
• /
• 한국항해항만학회 2006년도 Asia Navigation Conference
• /
• pp.105-114
• /
• 2006

Together with the trend of enhancement in domestic industrial development and economic progress due to import and export, the demand for construction of the roads, bridges, especially port facilities, and several coastal protection and ocean structures is increasing rapidly. MOMAF of Korean Government is driving construction of 9 new ports and renovation of the existing fishery ports. Among these structures most of bridge base, wharves, dolphins, quays, and jetties are being newly built of steel or concrete pile. As the base, supporting bulkheads, and piles are underwater after construction, it is difficult to figure out the status of structures and not enough to get maintenance and strengthen the structures. Every year, moreover, these works suck the government budget due to higher incomplete maintenance expense for protection from corrosions of structures and increased underwater construction period. for the purpose of cutting down the government budget, it is necessary to extend the life cycle of the existing structures. We developed a new method for maintenance of submerged structures near the waterline by allowing dry work environment with the floating caisson. The method shows easy to move around the working area and handle. It also showed not only a significant reduction of maintenance expenses and time for anti-corrosion work but also better protection. This will be a milestone to reduce the maintenance and construction expenses for the shore and water structures.

• 상하수도학회지
• /
• 제12권4호
• /
• pp.86-96
• /
• 1998

In this study, an intermittently aerated activated sludge process, modified process from conventional activated sludge process, was developed to treat high strength swine wastewater, which has been blamed as major pollutant for stream pollution. Therefore, the optimum cycle for oxic and anoxic period, SRT, and OLR were studied as design parameters. The effects of different time interval for oxic and anoxic period on nitrification and denitrification were examined by operating two reactors with 60/60min and 60/90min as oxic/anoxic period. Although the reactor with 60/60min showed complete denitrification of $NO_x-N$ generated during oxic period, the reactor with 60/90min showed incomplete nitrification due to the inactivity of nitrifier by accumulated $NH_3-N$ toxicity during anoxic period. Therefore, it is recommended to operate same interval for oxic and anoxic period. In order to determine the optimum cycle for oxic/anoxic period, four different reactors with 30/30, 60/60, 90/90 and 120/120min were examined. The reactor operation with 90/90min was optimum to get the most stable results in this study. However, the optimum cycle for oxic and anoxic period should be changed with characteristics of influent wastewater and operating conditions. According to lie operation results of three reactors with SRT of 15, 20 and 30days. The reactor with 2Odays SRT showed best removal efficiency of T-N. The optimum OLR would be $2.5Kg\;COD/m^3/day$ which showed the most stable nitrification and denitrification. Since characteristics of influent wastewater in the real system has a severe fluctuation, so it is very difficult to determine each interval for oxic and anoxic period. Therefore, ORP curves, describing the change of oxidation/reduction potential in reactor, can be used as a control parameter for automatic control of oxic and anoxic period. In other words, bending point (Nitrate Knee) of ORP curve during anoxic period could be used as a starting point of oxic period.

• 헤리티지:역사와 과학
• /
• 제43권3호
• /
• pp.180-207
• /
• 2010

21세기에 들어와서 나타난 급격한 기술의 진화는 사회 여러 분야에 큰 영향을 미치고 있다. 특히 컴퓨터의 발전은 다른 어느 분야보다도 급격한 진보를 이루었고 사회 전반에 걸쳐 새로운 변화와 효율성을 제고하고 있다. 고고학계도 예외는 아니어서 기존 수작업 방식의 현장조사 및 연구에서 탈피하려는 움직임이 활발히 일고 있다. 특히, 발굴되어진 불완전한 자료를 연구해야만 하는 고고학의 학문적 특성상 자료의 부족한 부분을 충족시켜줄 수 있는 방법의 하나로 컴퓨터를 활용하는 방법이 모색되고 있는데 그 중 중요한 부문의 하나로 GIS를 들 수 있다. 하지만 지금까지는 대부분 위치도나 유구배치도 같은 단순한 도면제작만을 위해 수치지도를 이용하고 있는 실정이며 GIS를 이용한 연구들도 방법적인 측면에 대해서는 자세히 설명하지 않고 있어 처음 GIS에 접근하는 연구자들은 그 활용법에 대해서 쉽게 익히기가 어렵다. 본고에서는 GIS를 처음 접근하는 연구자들도 손쉽게 익힐 수 있는 GIS의 기능에 대해서 소개하고 이를 위하여 먼저 GIS의 정의와 현재 고고학연구와 GIS가 어떻게 연관되어 있는지에 대해 소개하였으며 수치지도에서 GIS 공간분석에 필요한 DEM을 어떻게 추출하는지에 대해 설명하였다. 다음 장에서는 추출한 DEM을 이용하여 할 수 있는 GIS의 공간분석과 3D 시뮬레이션에 대해 소개 및 사용방법에 대해 언급하였다.