• Journal of Chest Surgery
• /
• 제27권9호
• /
• pp.752-758
• /
• 1994

From July 1983 to December 1992, 145 patients with mitral valvular disease underwent open heart surgery at Chonbuk National University Hospital. Of these patients, 89 patients[61.4%] required mitral valve replacement. 56 patients [38.6 %] had mitral valve repair. There were 32 women and 24 men and the mean age was 34.3 years[range 6 years to 62 years].There were 23 cases of pure mitral stenosis, 19 cases of mitral regurgitation and 14 cases of mixedmitral valvular disease. The mean duration of symptom was 4.53 years and mean mitral valvularorifice diameter[in cases of pure stenosis and mixed mitral valvular lesion] was 0.96 cm. According to the NYHA classification, the distribution of patients preoperatively was as follows; class IIa, 15 patients; class lib, 17 patients; class III, 22 patients; class IV, 2 patients. Four patients[7%] had an embolic history preoperatively. 24 patients[ 43 %] were in atrial fibrillation. In cases of pure mitral stenosis, the technique used included open mitral commissurotomy[21atients], open mitral commissurotomy with mitral annuloplasty[2 patients]. In mixed mitral valvular disease, open mitral commissurotomy[ll patients] and open mitral commissurotomy with mitral annuloplasty[l patient] were performed. In cases of mitral regurgitation, mitral annuloplasty[5 patients], mitral valvuloplasty[6 patients], mitral annuloplasty with valvuloplasty [3 patients] and ring annuloplasty [5 patients] were performed.There was one perioperative death related to acute renal failure and sepsis. One late death was occurred related to heart failure after 10 months postoperatively. One patient required reoperation due to restenosis and no embolic episode was occured. After operation, 34 patients were in NYHA functional class I, 20 patients were in class IIa.

• 한국정보처리학회논문지
• /
• 제7권8호
• /
• pp.2298-2309
• /
• 2000

본 논문에서는 객체지향 데이터베이스의 중포속성에 대한 색인기법으로 다차원 색인구조를 이용하는 다차원 중포속성 색인기법은 MD-NAI를 제안한다. 중포석성에 대한 기존의 색인기법들은 중포된 객체에 대한 기존의 색인기법들은 B+-tree와 같은 일차원 색인구조를 이용함으로써, 클래스 계층과 중포속성이 포함된 복합 형태의 질의들에 대한 처리를 잘 지원하지 못한다. MD-NAI에서는 객체지향 데이터베이스의 클래스 계층에 대한 색인기법인 이차원 클래스 계층 색인기법(2D-CHI)을 다차원으로 확장한다. 2D-CHI는 키 속성 도메인과 클래스 식별자 도메인으로 구성된 이차원 도메인 공간상에서 객체들의 클러스터링을 다루는 색인기법이다. 본 논문의 MD-NAI에서는 색인된 중포속성을 표현하는 경로상의 각 클래스 계층마다 하나의 클래스 식별자 도메인을 할당하여 구성된 다차원 도메인 공간상에서 색인 엔트리들의 클러스터링을 다룬다. 따라서, MD-NAI에서는 기존의 색인기법에서 지원하기 어려운 질의의 대상 범위 클래스 계층상의 임의의 클래스들로 제한되거나, 질의에 포함된 복합속성들의 도메인이 클래스 계층상의 임의의 클래스들로 제한되는 경우에도 잘 지원할 수 있다.

• 전자공학회논문지CI
• /
• 제47권3호
• /
• pp.22-28
• /
• 2010

최근 기계학습 분야에서 커널머신을 이용한 대표적 분류기로 Adaboost가 주목받고 있다. Adaboost는 통계적 학습이론에 기반하여 뛰어난 일반화 성능을 보여주며, 다양한 패턴인식 문제에 적용되고 있다. 그러나, Adaboost는 이진 분류기이므로 다중표적 분류 문제에 곧바로 적용할 수 없다. 일반적으로 다중 분류 문제를 해결하는 기법으로 One-Vs-All 기법과 Pair-Wise 기법이 대표적이다. 이러한 두 기법은 다중 분류 문제를 여러 개의 이진 분류 문제로 분할하고, 이들을 다시 종합하여 최종 결정을 내리는 출력코딩이라는 일반적인 기법으로 실제 시스템 구성에 적합할만한 분류 성능을 보여주지 못하는 경우가 대부분이다. 본 논문에서는 이진 분류기인 Adaboost의 다중 분류 확장 방안으로 원형 기반 함수를 약한 분류기로 이용하는 Adaboost 기반 다중표적 분류 기법을 제안한다.

• 한국생활과학회지
• /
• 제19권4호
• /
• pp.589-599
• /
• 2010

When the Korean government announced the lowering of the school entrance age from 6 to 5 years in an effort to increase the fertility rate, one candidate for the Seoul mayor rode on the slogan, 'one dong, one public kindergarten'. This study attempted to show the introduction of a preschool class for six-year-olds being used as analternative to lowering school entrance age in Sweden. First, the current state of preschool class, the Early Childhood Education and Care, and the education system in Sweden were described. Next, the background for the introduction of the preschool classes was analyzed. Thirdly, the effects of preschool class were evaluated from the perspectives of 6 year-olds as well as the institutiona land social status of the ECEC. Results from this study imply not only that the lower school entrance age policy should be abolished but also preschool classes could be considered as an alternative.

• 한국의류학회지
• /
• 제32권10호
• /
• pp.1595-1607
• /
• 2008

The purpose of this study is to find the reason why the National Mourning Dress System of the Joseon Dynasty had been established although the General Mourning Dress System had exist through research of three things; differences of each social class, changes of the National Mourning Dress System by the age and differences between the National one and the General one. The methodology which is adopted for this study is comparative analysis by social class, period and age from the Annals of the Joseon Dynasty and the Orye-e. And this study is focus on the women's dress. The results of this study is as follows; Firstly, some mourning dress items had been different or not used by class. Secondly, although there had not been a great change, some dress items had been disappeared or used irregularly. And Naemyoungbu Naekwan's mourning dress had been changed by the age. Thirdly, The National one was more extensive in wearing and applied differently by class compare to the General one. Also it had an additional rule according to the period, duty and place to Naemyoungbu-Naekwan and Naemyoungbu-Goongkwan. The National one had been established by ruling class of The Joseon Dynasty who had tried to keep the Confucian standard and perfect social status system. And it is supposed that the National Mourning Dress System had not been a great change because the ruling class' will had not been changed.

• 대한수학교육학회지:학교수학
• /
• 제4권3호
• /
• pp.331-346
• /
• 2002

This study has two purposes. One is to know that it is Possible to use portfolio in the elementary math class. The other is to make a useful method for using portfolio. We got the following conclusion through the study. Portfolio gave students an opportunity that they could review their mathematical thinking. But it couldn't work very well for the low-level students. They didn't pay attention to the class. So, careful prepa-ration and training were necessary for the portfolio material. And the portfolio material must be prepared by appropriate contents. Teacher had to do math class by considering students ability. The math class could be much better for motivation, teaching-learning activity impro-vement and communication tool by using portfolio material. There are several imple-mentation processes in preparation, execution and utilization of the class. 1) Preparation: Teacher must decide if it is appropriate for portfolio by analyzing the course and textbook and set a final goal. And then teacher has to select an appropriate item and make a schedule for the class. The portfolio material must contain valuable things from which students learn mathematics and use in their life. The student level, utilization purpose and contents are considered when one prepares portfolio material. 2) Execution: Students are supposed to understand about the portfolio very well. It is important for them to get the opportunity for reviewing through math class diary, their opinion, friends opinion and teachers opinion. 3) Utilization: Parents review ameliorates the communication among teacher, student and parents about learning activity.

• 한국IT서비스학회:학술대회논문집
• /
• 한국IT서비스학회 2008년도 추계학술대회
• /
• pp.157-160
• /
• 2008

The recent successes of systems like Google, Wikipedia and InnoCentive suggest that the time is now ripe for more such systems. Thus, we want to explore these systems and exploit them from a viewpoint of social psychology prior to technological progress. First of all, we examine the difference between 'wisdom of crowds' and 'crowd psychology'. There are two classes, which are for senior management major students. Conditions of two are same, except a class time. Forty five students in one class are tested for this experiment. They form a group with five people and are divided into 9 groups. In a case of a morning class, peer-to-peer evaluations are given to individuals in a class when a group presentation for a final project is given. On the other hand, in a case of afternoon class, peer-to-peer evaluations are given to each group in a class. The result is quite fruitful. The first is that an expert's evaluation for a project agrees with students, who are beginners in this field in the degree of more than 95%. However, in the case of afternoon class, students' group evaluation turns out to agree with an expert's one less than 53%. Morning class experiment result proves the truth of collective intelligence again. It tunrs out independent and confidential opinions are more cost effective and give right answers to questions than group opinions.

• 대한산업공학회지
• /
• 제38권3호
• /
• pp.191-197
• /
• 2012

One-class classification (OCC) is one of the recent growing areas in data mining and pattern recognition. In the present study we examine a k-nearest neighbors data description (kNNDD) algorithm, one of the OCC algorithms widely used. In particular, we propose to use nonparametric estimation methods to determine the threshold of the kNNDD algorithm. A simulation study has been conducted to explore the characteristics of the proposed approach and compare it with the existing approach that determines the threshold. The results demonstrate the usefulness and flexibility of the proposed approach.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제3권1호
• /
• pp.100-104
• /
• 2003

The SVDD(support vector data description) is one of the most well-known one-class support vector learning methods, in which one tries the strategy of utilizing balls defined on the kernel feature space in order to distinguish a set of normal data from all other possible abnormal objects. The major concern of this paper is to consider the problem of modifying the SVDD into the direction of utilizing ellipsoids instead of balls in order to enable better classification performance. After a brief review about the original SVDD method, this paper establishes a new method utilizing ellipsoids in feature space, and presents a solution in the form of SDP(semi-definite programming) which is an optimization problem based on linear matrix inequalities.

• 대한수학회보
• /
• 제52권5호
• /
• pp.1607-1619
• /
• 2015

We investigate some generalization of a class of Hankel-Clifford transformations having Fox H-function as part of its kernel on a class of Boehmians. The generalized transform is a one-to-one and onto mapping compatible with the classical transform. The inverse Hankel-Clifford transforms are also considered in the sense of Boehmians.