• 스마트미디어저널
• /
• 제4권4호
• /
• pp.18-23
• /
• 2015

본 논문에서는 텐서보팅을 이용하여 기존 마르코프 랜덤 필드 메소드 안의 가우시안 혼합 모델 함수의 성능을 향상시킬 수 있는 적합한 클러스터 개수 검출 방법을 제시한다. 제안하는 방법의 핵심 포인트는 텐서보팅의 인풋 데이터 토큰의 연속성인 saliency map을 통한 중심점 개수의 추출이다. 우리는 가장 먼저 주어진 자연 영상에서 전경 및 배경 후보 영역을 분리한다. 다음으로, 분리된 각 후보 영역에 대하여 텐서보팅을 적용하여 적절한 클러스터 개수를 추출한다. 우리는 검출된 클러스터 개수를 이용하여 정확한 가우시안 혼합 모델 모델링을 수행할 수 있다. 그리고 이를 적용한 마르코프 랜덤 필드의 unary term과 pairwise term을 계산하여 자연 영상의 텍스트 이진화 결과를 반환한다. 실험 결과, 제안된 방법이 최적의 클러스터 개수를 반환하고, 향상된 텍스트 이진화 결과를 반환함을 확인하였다.

• 한국정보과학회:학술대회논문집
• /
• 한국정보과학회 2005년도 가을 학술발표논문집 Vol.32 No.2 (1)
• /
• pp.214-216
• /
• 2005

We consider Wireless Sensor Network Security (WSN) and focus our attention to tolerate damage caused by an adversary who has compromised deployed sensor node to modify, block, or inject packets. We adopt a probabilistic secret sharing protocol where secrets shared between two sensor nodes are not exposed to any other nodes. Adapting to WSN characteristics, we incorporate these secrets to establish new pairwise key for node to node authentication and design multipath routing to multiple base stations to defend against HELLO flood attacks. We then analytically show that our defense mechanisms against HELLO flood attack can tolerate damage caused by an intruder.

• Asian-Australasian Journal of Animal Sciences
• /
• 제30권7호
• /
• pp.957-966
• /
• 2017

Objective: To understand the dynamic structure, function, and influence on nutrient metabolism in hosts, it was crucial to assess the genetic potential of gut microbial community in yaks of different ages. Methods: The denaturing gradient gel electrophoresis (DGGE) profiles and Illumina-based metagenomic sequencing on colon contents of 15 semi-domestic yaks were investigated. Unweighted pairwise grouping method with mathematical averages (UPGMA) clustering and principal component analysis (PCA) were used to analyze the DGGE fingerprint. The Illumina sequences were assembled, predicted to genes and functionally annotated, and then classified by querying protein sequences of the genes against the Kyoto encyclopedia of genes and genomes (KEGG) database. Results: Metagenomic sequencing showed that more than 85% of ribosomal RNA (rRNA) gene sequences belonged to the phylum Firmicutes and Bacteroidetes, indicating that the family Ruminococcaceae (46.5%), Rikenellaceae (11.3%), Lachnospiraceae (10.0%), and Bacteroidaceae (6.3%) were dominant gut microbes. Over 50% of non-rRNA gene sequences represented the metabolic pathways of amino acids (14.4%), proteins (12.3%), sugars (11.9%), nucleotides (6.8%), lipids (1.7%), xenobiotics (1.4%), coenzymes, and vitamins (3.6%). Gene functional classification showed that most of enzyme-coding genes were related to cellulose digestion and amino acids metabolic pathways. Conclusion: Yaks' age had a substantial effect on gut microbial composition. Comparative metagenomics of gut microbiota in 0.5-, 1.5-, and 2.5-year-old yaks revealed that the abundance of the class Clostridia, Bacteroidia, and Lentisphaeria, as well as the phylum Firmicutes, Bacteroidetes, Lentisphaerae, Tenericutes, and Cyanobacteria, varied more greatly during yaks' growth, especially in young animals (0.5 and 1.5 years old). Gut microbes, including Bacteroides, Clostridium, and Lentisphaeria, make a contribution to the energy metabolism and synthesis of amino acid, which are essential to the normal growth of yaks.

• 대한공업교육학회지
• /
• 제40권2호
• /
• pp.239-258
• /
• 2015

이 연구의 목적은 융합인재교육의 학습성과에 대한 합리적인 교육평가에 활용될 수 있는 기초자료를 제공하는 것으로, 융합인재교육에서 학습자가 함양해야 할 핵심 역량 준거의 상대적 가중치를 포함하는 학습성과 평가 방안을 제시하였다. 연구 방법은 융합인재교육 전문가를 대상으로 핵심역량에 대한 이원비교 설문조사를 실시하였고 핵심역량 요인의 우선순위를 분석하기 위하여 계층적 분석기법(AHP)을 적용하였다. 이에 따른 연구의 결론은 다음과 같다. 첫째, 핵심역량 1계층에서 융합수행능력과 융합인지능력의 중요도는 각각 39.4%, 36.8% 순으로 분석되었다. 융합인재교육의 교육평가에서 융합수행능력과 융합인지능력은 유사한 중요도 수준을 보였으며, 융합태도능력(23.8%)보다는 중요하게 고려해야 하는 요인으로 평가되었다. 둘째, 핵심역량 2계층에서 문제해결(20.0%)의 상대적 중요도가 가장 높게 나타났고, 다음으로는 창의적 사고력(18.3%), 책임감(15.3%), 융합지식 이해(11.0%) 순으로 분석되었다. 셋째, 핵심역량의 상대적 가중치 분석 결과는 융합인재 교육프로그램의 교육목표 달성 여부를 상대적 가중치에 따라 합리적으로 평가하는 역량평가 체제의 토대가 될 것이다. 더불어 융합인재교육의 교육평가뿐만 아니라 교수 학습 과정 개선과 학습효과를 높일 수 있는 교육프로그램 개발에 가이드 역할을 할 것으로 기대된다.

• ALGAE
• /
• 제27권4호
• /
• pp.225-234
• /
• 2012

Zygnema is a conjugating filamentous green algal genus that is distributed in a broad range of freshwater habitats, from sea level to alpine summits. Although more than 150 species have been described worldwide, their taxonomy remains unclear, probably owing to their relatively simple morphology. We investigated the detailed morphology of Korean Zygnema species, combined with analysis of the plastid psbA gene from 22 specimens of the genus and putative relatives, in order to develope a key to their identification and isolation, and to determine their relationships. We recognized two species of Zygnema; Z. insigne and Z. leiospermum, based on morphological characters such as width of the vegetative cell, position of zygospores, dimensions and form of spores, shape of female gametangia, and color of mesospores. The analysis of psbA data was consistent with morphological comparison. The pairwise divergence between two species was 3.7-4.1% (34-38 bp) in psbA sequences. The phylogeny of psbA revealed the monophyly of Z. insigne and Z. leiospermum together with two isolates of Z. circumcarinatum from Germany and Scotland. This is the first report on the psbA gene phylogeny of Zygnema.

• Journal of Communications and Networks
• /
• 제5권2호
• /
• pp.141-149
• /
• 2003

In this paper, we investigate design criteria and the performance of the space-frequency bit-interleaved coded modulation (SF-BICM) systems in frequency-selective Rayleigh fading channels. To determine the key parameters that affect the performance of SF-BICM, we derive the pairwise error probability (PEP) in terms of the determinant of the matrix corresponding to any two codewords. We prove that the bit-interleavers do the function of distributing the nonzero bits uniformly such that two or more nonzero bits are seldom distributed into the symbols that are transmitted in the same frequency bin. This implies that the bit-interleavers transform an SF-BICM system into an equivalent 1-antenna system. Based on this, we present design criteria of SFBICM systems that maximizes the diversity order and the coding gain. Then, we analyze the performance of SF-BICM for the case of 2-transmit antennas and 2-multipaths by deriving a frame error rate (FER) bound. The derived bound is accurate and requires only the distance spectrum of the constituent codes of SF-BICM. Numerical results reveal that the bound is tight enough to estimate the performance of SF-BICM very accurately.

• Communications for Statistical Applications and Methods
• /
• 제27권3호
• /
• 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.

• 한국동물생명공학회지
• /
• 제36권4호
• /
• pp.314-322
• /
• 2021

Paired box protein, PAX7, is a key molecule for the specification, maintenance and skeletal muscle regeneration of muscle satellite cells. In this study, we identified and characterized the cDNA and amino acid sequences of PAX7 from black sea bream (Acanthopagrus schlegelii) via molecular cloning and sequence analysis. A. schlegelii PAX7 cDNA was comprised of 1,524 bp encoding 507 amino acids and multiple sequence alignment analysis of the translated amino acids showed that it contained three domains including paired DNA-binding domain, homeobox domain and OAR domain which were well conserved across various animal species investigated. Pairwise Sequence Alignment indicated that A. schlegelii PAX7 had the same amino acid sequences with that of yellowfin seabream (A. latus) and 99.8% identity and similarity with that of gilt-head bream (Sparus aurata). Molecular phylogenetic analysis confirmed that A. schlegelii PAX7 formed a monophyletic group with those of teleost and most closely related with those of the fish that belong to Sparidae family including A. latus and S. aurata. In the investigation of its tissue specific mRNA expression, the expression was specifically identified in skeletal muscle tissue and a weak expression was also shown in gonad tissue. The cultured cells derived from skeletal muscle tissues expressed PAX7 mRNA at early passage but the expression was not observed after several times of subculture.

• 한국콘텐츠학회논문지
• /
• 제14권4호
• /
• pp.407-417
• /
• 2014

우리 민족의 대표적인 민요이면서 동시에 유네스코 인류무형문화유산인 아리랑을 정보알고리즘 기법을 도입하여 후렴구를 중심으로 계통도를 분석하고 아리랑들 사이의 상관관계는 본문 단어중심으로 분석하였다. 아리랑의 계통도 분석은 생명체의 진화관계를 분석하는 알고리즘인 다중서열정렬 기법을 사용하였다. 분석한 아리랑 106개 중에서 38개 아리랑이 빠른 템포를 가지고 있었으며, 나머지 68개 아리랑이 느린 템포를 가지고 있었다. 이를 바탕으로 후렴구 기반 아리랑 계통도를 완성하였다. 아리랑 본문 단어는 아리랑에 있는 단어와 아리랑 제목을 노드로 하는 bipartate네트워크를 구축하고 이들로부터 73개 아리랑 및 104개의 핵심 단어를 추출하였다. 먼저, 이 데이터를 바탕으로 쌍대비교분석 기법을 사용하여 아리랑들 사이의 상관관계를 분석하였다. 또한, 네트워크 연결계수가 1인 노드를 단계적으로 제거하여 핵심네트워크를 구축한 다음 네트워크 기반으로 아리랑들 사이의 상관관계를 분석하였다. 그동안 아리랑을 어원 중심의 인문과학이나 음률적인 접근을 통하여 아리랑의 어원, 계통도, 상관관계를 분석하려는 연구가 있었다. 본 연구에서는 이러한 시도를 벗어나 과학적 접근방법인 정보알고리즘을 사용하여 아리랑을 분석함으로써 세계적인 문화유산의 위상을 한층 더 높이고 객관적인 결과를 통해서 아리랑의 대중화 및 세계화의 기틀을 마련함에 있어 그 방법론을 제시하였다.

• 한국병원경영학회지
• /
• 제22권1호
• /
• pp.40-50
• /
• 2017

The purpose of this study is to develop the KPIs(Key Performance Indices) needed to improve management and strategy in the dental clinic based on the four perspectives of BSC(Balanced Scorecard). The questionnaire was conducted on 52 dentists approved by Dental Managment Research Committee in Seoul National University as a panel. Using the Delphi technique, the top five KPIs for each point of perspective in BSC were extracted from KPI pools. In the third survey, the top five KPIs of all points were compared with each other through AHP(Analytic Hierarchy Process) method, and priority and overall importance rankings were calculated. The biggest difference in the three level AHP results was the customer perspective took priority to others. In the second survey, the financial perspective, which was number one, was pushed back. The overall significance of KPIs was in the order of customer, internal process, finance, learning and growth perspective, with the exception of medical profits (5th of 20) and new patient growth (10th of 20). We were able to overcome the limitations of the Delphi Technique with the AHP method. In general, the financial perspective in BSC is known to be the most important, but we conclude that the customer perspective is more important through the pairwise comparison survey. In the current dental service market, which is a long-term recession, excessive competition, customer satisfaction and customer relationship management seem to be the first goal to pursue in dental clinic.