• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제16권1호
• /
• pp.60-79
• /
• 2022

The U-Net architecture-based segmentation models attained remarkable performance in numerous medical image segmentation missions like skin lesion segmentation. Nevertheless, the resolution gradually decreases and the loss of spatial information increases with deeper network. The fusion of adjacent layers is not enough to make up for the lost spatial information, thus resulting in errors of segmentation boundary so as to decline the accuracy of segmentation. To tackle the issue, we propose a new deep learning-based segmentation model. In the decoding stage, the feature channels of each decoding unit are concatenated with all the feature channels of the upper coding unit. Which is done in order to ensure the segmentation effect by integrating spatial and semantic information, and promotes the robustness and generalization of our model by combining the atrous spatial pyramid pooling (ASPP) module and channel attention module (CAM). Extensive experiments on ISIC2016 and ISIC2017 common datasets proved that our model implements well and outperforms compared segmentation models for skin lesion segmentation.

• 한국대기환경학회지
• /
• 제16권2호
• /
• pp.141-150
• /
• 2000

In order to genralize the vertical dispersion of plume at long distances on mesoscale over complex terrain dispersion coefficients data have been obtained systematically according to lapsed time after release by using a composite turbulence water tank that simulates convective boundary layer. Dispersion experiments have been carried out for various combined conditions of thermal turbulence intensity mechanical turbulence intensity and plume release height at slightly to moderately unstable conditions. Results of tracer dispersion experiments conducted using water tank camera and image processing system have been converted into atmospheric dispersion data through the application of similarity law. The equation $\sigma$z/Zi=aX/(b+c X2)0.5 where $\sigma$2; vertical dispersion coefficient zi : mixing height X : dimen-sionaless downwind distance was confirmed to be an appropriate and general equation for expressing $\sigma$2 variation with turbulence intensity and plume release height, The value of "a" was found to be principally affected by mechanical turbulence intensity and that of "b" by mechanical turbulence intensity and release height. It was confirmed that the magnitude of "c" varies with release height. Results of water tank experiments on the relationship of $\sigma$2 vs downwind distance x have been compared with actual atmospheric dispersion data such as CONDORS data and Bowne's nomogram Operating conditions of a composite turbulence water tank for simulating the field turbulence situations of CONDORS experiments and Bowne's $\sigma$2(x) nomogram for suburban area have also been investigated in terms of water temperature difference between convection water tank and bottom plate heating tank grid plate stroke mixing water depth length scale and velocity scale. Moreover the effect of mechanical turbulence intensity on vertical dispersion has been discussed in the light of release height and downwind distance. height and downwind distance.

• 전자공학회논문지CI
• /
• 제49권2호
• /
• pp.53-60
• /
• 2012

최근접 이웃 분류에서 입력 데이터의 클래스는 선택된 근접 학습 데이터들 중에서 가장 빈번한 클래스로 예측된다. 최근접분류 학습은 학습 단계가 없으나, 준비된 데이터가 모두 예측 분류에 참여하여 일반화 성능이 학습 데이터의 질에 의존된다. 그러므로 학습 데이터가 많아지면 높은 기억 장치 용량과 예측 분류 시 높은 계산 시간이 요구된다. 본 논문에서는 분리 경계면에 위치한 학습 데이터들로 구성된 새로운 학습 데이터를 생성시켜 분류 예측을 수행하는 프로토타입 선택 알고리즘을 제안한다. 제안하는 알고리즘에서는 분리 경계 영역에 위치한 데이터를 Tomek links와 거리를 이용하여 선별하며, 이미 선택된 데이터와 클래스와 거리 관계 분석을 이용하여 프로토타입 집합에 추가 여부를 결정한다. 실험에서 선택된 프로토타입의 수는 원래 학습 데이터에 비해 적은 수의 데이터 집합이 되어 최근접 분류의 적용 시 기억장소의 축소와 빠른 예측 시간을 제공할수 있다.

• 대한지리학회지
• /
• 제39권5호
• /
• pp.804-818
• /
• 2004

공간데이터자원이 많아 질수록 그들끼리 불일치가 일어날 확률은 높아지고 있다. 이러한 불일치는 같은 지역을 커버하는 같은 종류의 공간데이터사이에서도 일어날 수 있다. 그러므로, 이런 공간데이터를 효율적으로 연결시켜 Geometry 및 Topology측면에서 일관성을 지닌 새로운 공간데이터를 생성시키는 일의 중요성은 증가 할 것이다. 이러한 공간데이터중의 하나로서 미국 인구통계국의 TIGER파일을 예로 들 수 있다. 현재 인구통계국 지도들은 지방정부의 지도 레이어들과 공간적으로 일치 하지 않기 때문에 인구적, 경제적인 많은 유용한 정보가 지방정부의 레이어들과 연계되어 공간적으로 충분히 활용되어지고 있지 않고 있다. 그러므로, 인구통계국 지도의 위치정보는 좀 더 정확한 위치정보를 가지고 있는 지방정부의 레이어들과 융합되어 Geometry 및 Topology측면에서 새로운 정보로 대체되어져야 한다. 이 논문은 참고맵을 이용하여 Geometry 및 Topology측면에서 일관성을 지닌지도를 만들기 위한 개념적인 프레임과 두가지 맵모델을 제시한다. 첫번째 모델은 셀 모델인데 맵은 0셀, 1셀, 그리고 2셀로 구성되어진다. 두번째 모델은 수학적으로 다른 원형을 가진 물체는 지도 일반화후에도 유사성을 가지고 있다는 것이다. 새롭게 제시된 계층적인 맵 융합은 물리적, 수학적, 논리적 경계에 바탕을 두고 있고 복잡성과 계산적인 부담을 감소시킬 수 있다. 반복성을 가진 맵 융합 원리는 인구통계지도를 예로하여 형성되었다. 이것들은 속성 매치, 의미있는 노드발견, 지도학적인 0-cell 매치. 지도학적인 1-cell 매치, 그리고 맵 변형으로 구성된다.

• 한국언어정보학회지:언어와정보
• /
• 제18권1호
• /
• pp.33-52
• /
• 2014

This paper proposes a unified structural account of some instances of the English Modals and Semi-auxiliaries. The classification and the syntactic/structural description of the English Modal auxiliary verbs and verb-related elements have long been the center for many proposals in the history of generative syntax. According to van Gelderen (1993) and Lightfoot (2002), it was sometime around 1380 that the Tense-node (T) appeared in the phrasal structures of the English language, and the T-node is under which the English Modal auxiliaries occupy. Closely related is the existing evidence that English Modals were used as main verbs up to the early sixteenth century (Lightfoot 1991, Han 2000). This paper argues for a bi-clausal approach to English Modal auxiliaries with the infinitival particle 'to' such as 'ought to' 'used to' and 'dare (to)' 'need (to)', etc. and Semi-auxiliaries including 'be to' and 'have to'. More specifically, 'ought' in 'ought to' constructions, for instance, undergoes V-to-T movement within the matrix clause, just like 'HAVEAux' and all instances of 'BE', whereas 'to' occupies the T position of the embedded complement clause. By proposing the bi-clausal account, Radford's (2004, 2009) problems can be solved. Further, the historical motivation for the account takes a stance along with Norde (2009) and Brinton & Traugott (2005) in that Radford's (2004, 2009) syncretization of the two positions of the infinitival particle 'to' is no different from the 'boundary loss' in the process of Grammariticalization. This line of argument supports Krug's (2011), and in turn Bolinger's(1980) generalization on Auxiliaryhood, while providing a novel insight into Head movement of V-to-T in Present Day English.

• 한국정보처리학회논문지
• /
• 제3권2호
• /
• pp.381-389
• /
• 1996

본 논문은 신경망에서 불필요한 연결자(weights and biases)를 제거하기 위한 일반적인 back-propagation 알고리즘의 간단한 변형을 소개한다. 이는 연결자들의 절대치 크기의 분포, 분할 경계선의 분석 및 sigmoid 함수의 비선형성에 기초하여 개발되었다. 신경망의 초기 학습 후, 이 알고리즘은 그 절대치의 크기가 임계치 (threshold) 보다 적은 모든 연결자들의 값을 영으로 할당 함으로서 이들을 제거하게 된다. 그런후, 중요한 연결자들의 복구를 위해 모든 연결자들을 포함하여 반복 학습을 실시한다. 이진함수 들을 이용한 학습에서, 이 알고리즘은 이론적 최소 구조를 실현하 였고, 함수를 푸는데 불필요한 연결자들을 제거하였다. 모의 랜덤 신호 분류에 응용에 있어서, 본 알고리즘으로 부터 얻어낸 결과는 쉬운 문제가 간단한 신경망을 필요로 하며 낮은 오분류율을 발생한다는 일반적인 개념과 일치 하였다. 또한, 본 제안된 알고리즘 은 overfitting과 형태소 (patterm) 암기의 문제점들을 줄임으로서 기존의 알고리즘보 다 더 좋은 결과를 보여주었다.

• 영미문화
• /
• 제9권1호
• /
• pp.239-272
• /
• 2009

Our investigation of the intricate relationship among nationalism, humanism, and imperialism begins from reading Song of Ariran, the auto/biography of Kim San recorded by Nym Wales, together with Margaret Drabble's fictional adaptation of Lady Hong's autobiography, The Memoirs of Lady $Hyegy{\breve{o}}ng$, in her novel The Red Queen, in which the story of Barbara Halliwell, a modern female envoy of Lady Hong, is interweaved with Lady Hong's narrative. In spite of their being seemingly disparate texts, Song of Ariran and The Red Queen are comparable: they are written by Western female writers who deal with Koreans, along with the Korean history and culture. Accordingly, both works cut across the boundary of fiction and fact, imagination and history, and the East and the West. In the age of globalization, Western women writing (about) Korea and Koreans traversing the historical and cultural limits inevitably engage us in post-colonial discussions. Despite the temporal differences--If Song of Ariran handles with the historical turmoils of the 1930s Asia, mostly surrounding Kim San's activities as a nationalist, The Red Queen is written by a twenty-first century British woman writer whose international interest grapples with the eighteenth-century Korean Crown Princess' spirit in order to reinscribe a story of Korean woman's within the contemporary culture--, both works appeal to the humanistic perspective, advocating the universal human beings' values transcending the historical and national limitations. While this sort of humanistic approach can provide sympathy transcending time and space, this 'idealistic' process can be problematic because the Western writers's appropriation of Korean culture and its history can easily reduce its particularities to comprehensive generalization, without giving proper names to the Korean history and culture. Nonetheless, the Western female writers' attempt to find a place of 'contact' is valuable since it opens a possibility of having meaningful communications between minor culture and dominating culture. Yet, these female writers do not seem to absolutely cross the border of race, gender, and culture, which leaves us to realize how difficult it is to reach a genuine understanding with what is different from mine even in these 'universal' narratives.

• 대한음성학회:학술대회논문집
• /
• 대한음성학회 1996년도 10월 학술대회지
• /
• pp.109-116
• /
• 1996

In this paper, I would like to explore the possibility that the nature of place assimilation can be captured in terms of the OCP within the Optimality Theory (Mccarthy & Prince 1999. 1995; Prince & Smolensky 1993). In derivational models, each assimilatory process would be expressed through a different autosegmental rule. However, what any such model misses is a clear generalization that all of those processes have the effect of avoiding a configuration in which two consonantal place nodes are adjacent across a syllable boundary, as illustrated in (1):(equation omitted) In a derivational model, it is a coincidence that across languages there are changes that have the result of modifying a structure of the form (1a) into the other structure that does not have adjacent consonantal place nodes (1b). OT allows us to express this effect through a constraint given in (2) that forbids adjacent place nodes: (2) OCP(PL): Adjacent place nodes are prohibited. At this point, then, a question arises as to how consonantal and vocalic place nodes are formally distinguished in the output for the purpose of applying the OCP(PL). Besides, the OCP(PL) would affect equally complex onsets and codas as well as coda-onset clusters in languages that have them such as English. To remedy this problem, following Mccarthy (1994), I assume that the canonical markedness constraint is a prohibition defined over no more than two segments, $\alpha$ and $\beta$: that is, $^{*}\{{\alpha, {\;}{\beta{\}$ with appropriate conditions imposed on $\alpha$ and $\beta$. I propose the OCP(PL) again in the following format (3) OCP(PL) (table omitted) $\alpha$ and $\beta$ are the target and the trigger of place assimilation, respectively. The '*' is a reminder that, in this format, constraints specify negative targets or prohibited configurations. Any structure matching the specifications is in violation of this constraint. Now, in correspondence terms, the meaning of the OCP(PL) is this: the constraint is violated if a consonantal place $\alpha$ is immediately followed by a consonantal place $\bebt$ in surface. One advantage of this format is that the OCP(PL) would also be invoked in dealing with place assimilation within complex coda (e.g., sink [si(equation omitted)k]): we can make the constraint scan the consonantal clusters only, excluding any intervening vowels. Finally, the onset clusters typically do not undergo place assimilation. I propose that the onsets be protected by certain constraint which ensures that the coda, not the onset loses the place feature.

• 한국식생활문화학회지
• /
• 제22권6호
• /
• pp.653-664
• /
• 2007

This study intends to consider about the food culture during the latterly Chosun Dynasty through the genre paintings which are depicted realistically about the total food life figures for example not only the foods, but also that people who eat and drink the food and eating implements, eating methods. There are 9 pieces of works Kim hong-do's which are the best appraisal on the incorporated Korean emotion and on the real life of general public. Those are selected for this study's characteristics through the literature consideration. The characteristics of the tableware pottery and food culture are following things. The first is a white porcelain's development and popularization. In the latterly Chosun Dynasty period, the development of the agriculture and commerce made been connected to the affluent food cultures, these made increase the porcelain's use of upper classes and these made been connected to the development of the porcelain's cultures. The second is the development of various pottery with a dark brown glaze manufactures. The development of agriculture and economy power made bring the development of the kinds of sauces and Kimchi culture with an inflow of foreign crops and these made making the large an earthenware pot or jar which could store something during the long time. The third is the development of the small dining table on the eating the bread of idleness. In the Chosun dynasty period, the food style was been unified to he eating the bread of idleness and been lowered. These made developed of various small dining table through the underfloor heating system's generalization. The fourth is the development of the spoon with the soup culture. It is the East Asia called that the spoon culture boundary, this show the characteristic that we use the spoon mainly compare with another countries. Finally, it is universal use of the fixed quantitative implements. They could live quantitative life with the scales which are jar of soy sauce, liquor jug, jar, scoop leading a set of tableware.

• 지능정보연구
• /
• 제18권2호
• /
• pp.29-45
• /
• 2012

회사채 신용등급은 투자자의 입장에서는 수익률 결정의 중요한 요소이며 기업의 입장에서는 자본비용 및 기업 가치와 관련된 중요한 재무의사결정사항으로 정교한 신용등급 예측 모형의 개발은 재무 및 회계 분야에서 오랫동안 전통적인 연구 주제가 되어왔다. 그러나, 회사채 신용등급 예측 모형의 성과와 관련된 가장 중요한 문제는 등급별 데이터의 불균형 문제이다. 예측 문제에 있어서 데이터 불균형(Data imbalance) 은 사용되는 표본이 특정 범주에 편중되었을 때 나타난다. 데이터 불균형이 심화됨에 따라 범주 사이의 분류경계영역이 왜곡되므로 분류자의 학습성과가 저하되게 된다. 본 연구에서는 데이터 불균형 문제가 존재하는 다분류 문제를 효과적으로 해결하기 위한 다분류 기하평균 부스팅 기법 (Multiclass Geometric Mean-based Boosting MGM-Boost)을 제안하고자 한다. MGM-Boost 알고리즘은 부스팅 알고리즘에 기하평균 개념을 도입한 것으로 오분류된 표본에 대한 학습을 강화할 수 있으며 불균형 분포를 보이는 각 범주의 예측정확도를 동시에 고려한 학습이 가능하다는 장점이 있다. 회사채 신용등급 예측문제를 활용하여 MGM-Boost의 성과를 검증한 결과 SVM 및 AdaBoost 기법과 비교하여 통계적으로 유의적인 성과개선 효과를 보여주었으며 데이터 불균형 하에서도 벤치마킹 모형과 비교하여 견고한 학습성과를 나타냈다.