• 방송공학회논문지
• /
• 제10권2호
• /
• pp.149-155
• /
• 2005

We have proposed an objective evaluation method using ESNR as the measure of approximation by the visual model, which coincides with MOS, a subjective evaluation method. For two-value images, we have used five kinds of modulation methods: 1) ordered dither, 2) least mean error, 3) pulse density low division, 4) simple two-value, and 5) random dither methods. The purpose of this paper is to investigate the validity of ESNR, by comparing the proposed method together with the existing representative methods such as PSNR and WSNR, with the subjective method MOS. The results of a series of experiments show that the ranking by MOS coincides with ESNR, though does not coincides with PSNR and WSNR.

• 한국경영과학회지
• /
• 제16권1호
• /
• pp.13-34
• /
• 1991

A production/inventory system is considered in which a production facility produces one type of product. The demand for the product is given by a compound Poison process and is supplied directly from inventory when inventory is available and is lost when inventory is out of stock. The processing time to produce one item is assumes to follow a general distribution. An (s, S) policy is considered in which production stops at the instant the stock on hand reachs S and the setup of the production facility begins at an inspection point when the stock on hand drops to or below s for the first time. The time interval between two successive inspection points during a non-production period is a random variable which follows a general distribution.

• 산업경영시스템학회지
• /
• 제22권50호
• /
• pp.1-9
• /
• 1999

The purpose of this study is to establish a basis for assessing construct validity of measures used in organizational research. The classic Campbell and Fiske's(1959) criteria are found to be lacking in their assumptions, diagnostic information, and power. The inherent confounding of measurement error with systematic trait and method effects is a severe limitation for a proper interpretation of convergent and discriminant validity. The confirmatory factor analysis(CFA) approach overcomes most of the limitations found in Campbell and Fiske's(1959) method. However, the CFA approach confounds random error with unique variance specific to a measure. The second-order confirmatory factor analysis(SOCFA) was shown to harbor rather restrictive assumptions and is unlikely to be met in practice. The first-order, multiple-informant, multiple-item(FOMIMI) model is a viable option, but it may also be of limited use because of the large number measures.

• 대한산업공학회지
• /
• 제33권3호
• /
• pp.322-328
• /
• 2007

Highly standardized products are suitable for automated purchasing using electronic commerce technology, where the price becomes the most important factor. Suppliers can change the prices dynamically based on the inventory level and market situation in order to maximize the sales and profit. In the virtual marketplace where multiple customers purchase multiple standardized products from multiple suppliers repetitively, customers can purchase the required amount of each item as a dynamic bidding by allocating purchase orders to the suppliers based on the current price. Customers need a method to quickly determine the optimal allocation of orders to the suppliers using the dynamically changing data to minimize the total cost. We present a LP model which minimizes the sum of the total price plus transportation cost for this problem. Simulation results using random data show meaningful reduction of the total cost.

• 대한물리의학회지
• /
• 제6권4호
• /
• pp.425-436
• /
• 2011

Purpose : The purpose of this study is to identify static, dynamic balance performances and foot pressure after eye movement exercise in normal adults. Method : This study was performed on normal 18 males and 18 females subjects. They were divided into 3 groups using random sampling and executed saccadic eye movement, vestibuloocular eye movement, pursuit vergence eye movement exercise respectively. In order to compare the difference of balance and foot pressure, the subjects were measured before, middle and after eye movement. Balance was examined using the balance master 7.0 version systems. Foot pressure was examined using the gaitview AFA-50. Result : The result are as follows. 1. After first intervention, the difference of the FORM-EC item on MCTSIB variable in the vestibuloocular group was significantly decreased(p<.05) and after second intervention, the difference of the FIRM-EO(p<.05), FORM-EO(p<.01), FORM-EC(p<.01) items on MCTSIB variable in the saccadic group and FORM-EC(p<.05) item in the vestibuloocular group were significantly decreased respectively. 2. After first intervention, the difference of the EPE, MXE items in the saccadic group(p<.05) and RT(p<.05), EPE(p<.01) items in the pursuit vergence group were significantly decreased respectively after second intervention. 3. After first intervention, the difference of the Static A-P(p<.05), L-R(p<.01) and Dynamic A-P(p<.01), L-R (p<.01) items on foot pressure variable in the saccadic group, Dynamic A-P(p<.01) item in the vestibuloocular group and Static A-P(p<.01), Dynamic A-P(p<.05), L-R(p<.01) items in the pursuit vergence group were significantly decreased respectively. 4. After first intervention, the difference of the FORM-EO(p<.05), FORM-EC(p<.01) items and after second intervention, the difference of the FIRM-EO(p<.01) item on MCTSIB variable among three groups was significantly decreased respectively. Conclusion : Selective eye movement exercise program influences balance control ability and foot pressure of normal adults in positive ways.

• 한국정보통신학회:학술대회논문집
• /
• 한국해양정보통신학회 2009년도 추계학술대회
• /
• pp.599-601
• /
• 2009

현재 이러닝 표준으로 채택하고 있는 SCORM은 런타임 시 학습자의 수준 변화에 따른 맞춤형 콘텐츠 제공이 어렵고, 선택적 학습을 제어하기 어렵다. 따라서 본 논문에서는 이러한 SCORM의 단점을 보완하기 위하여, 수준 평가 모듈, 콘텐츠 재구성 모듈, 문항 출제 모듈로 구성된 맞춤형 콘텐츠 구성 엔진(CCOE : Customized Contents Organization Engine)을 설계 및 구현하였다. 수준 평가 모듈은 문항반응이론을 기반으로 학습자의 수준을 평가하고, 문항 출제 모듈은 각 수준별로 랜덤하게 또는 학습자의 수준에 적합한 문항들을 추출하여 학습 이전 평가, 단원 평가 및 퀴즈로 제공하며, 퀴즈로 제공하기 위해 추출된 문항들을 콘텐츠 재구성 모듈로 전달한다. 콘텐츠 재구성 모듈은 콘텐츠에 태깅된 난이도를 검색하여 학습자의 수준에 적합한 콘텐츠를 추출하고, 문항 출제 모듈로부터 전달받은 퀴즈와 추출된 콘텐츠에 대한 시퀀스를 생성한다. 본 논문에서 제안한 CCOE를 활용하면, 각 단원별로 변화된 학습자의 수준을 재평가하여 변화된 수준에 적합한 학습 콘텐츠를 제공함으로써 학습 효과를 더 높일 수 있을 것으로 기대된다.

• 가정과삶의질연구
• /
• 제26권6호
• /
• pp.209-225
• /
• 2008

This study has been designed to develop integrated supportive service works for the families of disabled children registered in the D-Healthy Family Support Center. The service works are intended to improve family members' health and their growth potentials, and cover service programs of family education, counselling, culture solidarity, and information networks. There are two topics covered in this study; first, this study is to develop integrated supportive service programs for the families of disabled children, and second, the study is also to construct comprehensive strategies and plans for effective services after evaluating practical service programs. In this study, 31 families with one or more disabled children have been selected through stratified random sampling, and all the family members were observed for program effects of integrated service items developed from previously validated indices. For analysis, eight question items were made for interview and each item denotes satisfaction score ranging from 1(very dissatisfied) to 5(very satisfied) on the Likert's five point scale. Mean value of each item was counted after the program, and satisfaction scores were compared over the programs implementation. Overall, the satisfaction scores range 4.1 to 4.6, indicating relatively high satisfaction over most service program. Therefore, it is finally proposed that a longitudinal strategy model offer high quality service programs consistently for the target families, as a goal of achieving comprehensive cooperation and support exchanges between community and healthy family support center be schematized.

• 복식
• /
• 제50권2호
• /
• pp.97-109
• /
• 2000

The purpose of this study is to apply 3D computer graphics in fashion design as a creative medium and it attempts to fine out what advantages 3D technique can offer to fashion design. For this purpose, this study, first, tries to develop a 3D digital model in which designer can select design, color , pattern and fabric palette whatever necessary . This study uses of the software named 'Poser of Fractal Design' and the and the 3D digital model comprises four stages ; body modeling, item design (item coordination), color design (color coordination), pattern and fabric design (pattern and fabric coordination). Secondly, this study seeks to accumulate a data base which was produced in the course of case studies, which have applied 3D digital model to design. The outcome of the case studies shows that 3D digital model can enhance designing in the following four aspects. ⅰ) It can give more freedom to designer to try various ideas, revise and modify them, It can also produce random generation. Through this process, the designer test various input and output without damage on fabric after revision and alteration. ⅱ) It can help designers to enhance their accuracy. Since fault in the design developed by the 3D digital model can be detected in advance, designers can make correlation before actual work begins. In the end, designers can express their ideas and intention accurately as well as freely. ⅲ) Since design developed by the 3D digital model can be shared on screen by various actors in the course of priduction such designers , merchandisers, and supervisors, it can help communication between and cut the time of feedback.ⅳ) By using the 3D digital model, designers can work from the begining with awareness of actual outcome their design, since the 3D digital model provide animation, which helps designers to envisage visual changes as they apply various items, colors, pattern and fabrics.

• 한국산업정보학회논문지
• /
• 제28권6호
• /
• pp.11-20
• /
• 2023

자동차보험 교통사고 진료비는 매년 증가하고 있다. 본 연구는 교통사고 진료비용 상승의 주요 항목인 경상환자 중 장기입원환자(18일 이상)를 예측하는 모델을 decision tree 등 5개 알고리즘을 이용하여 생성하고, 장기입원에 영향을 미치는 요인을 분석했다. 그 결과, 예측 모델의 정확도는 91.377 ~ 91.451이며 각 모델 사이에 큰 차이점은 없었으나 random forest와 XGBoost 모델이 91.451로 가장 높았다. 설명변수 중요도에 있어서 병원 소재지, 상병명, 병원 종류 등 장기환자군과 비 장기입원 환자군 사이에 모델마다 상당한 차이가 있었다. 모델 평가는 훈련 데이터의 교차검증(10회)한 모델별 평균 정확도와 실험 데이터의 정확도를 상호 비교한 결과로 검정했다. 설명변수 유의성 검증을 위해 범주형 변수는 카이제곱 테스트를 실시하였다. 본 논문의 연구 결과는 경상 환자들의 과잉진료 및 사회적 보험료 비용을 줄이는 진료행태 분석에 도움이 될 것이다.

• Journal of the Korean Academy of Child and Adolescent Psychiatry
• /
• 제9권1호
• /
• pp.91-97
• /
• 1998

저자들이 번안한 한국판 청소년용 식이태도 검사(Eating Attitude Test for Korean Adolescents, EAT-26KA)의 신뢰도 조사와 요인분석을 통한 개개의 문항에 대한 평가를 하고자 하였다. 경기도 소재의 1개 중 고등학교 학생 1,160명을 대상으로 하여 방문 조사 2주전 양호교사와 담임교사를 통해 설문지를 배포하고 1996년 6월 저자들이 직접 현장 방문을 하여 신체계측과 누락된 문항을 보강하는 검사를 하고 1개월후 이중 114명을 무작위 선정하여 동일한 방법으로 재검사를 시행하였다. 평가도구인 EAT-26KA는 영문판 EAT-26의 전 문항을 사용하였고 저자들이 대상연령의 지적 능력에 맞는 어휘를 사용하여 번역하였다. 내적 일관성 혹은 내적 합치도는 우수한 것으로 나타났다(Cronbach’s ${\alpha} =.83) 검사-재검사 신뢰도는 서구의 조사결과에 비해선 약간 낮게 나타났으나 통계적으로 유의한 수준이었고(r=.63 p<.01) 문항간 검사-재검사 신뢰도에선 두 개의 문항이 부적 상관(negative correlation)을 보였고 다섯 문항에서 문항간 상관이 낮게 나타났다(p>.05). 세 가지 요인으로 분류하여 요인분석을 시행한 결과 세 요인은 전체요인의 34.0%를 설명하였고 한 문항은 어느 요인으로도 분류되지 않았다. 각 요인으로 분류되는 문항구성도 영문판에 비해서 많은 차이를 보였다. 저자들이 번안한 EAT-26KA는 일반 청소년 인구(non-refered population)을 통한 신뢰도 검사 결과 우수한 내적 신뢰도를 갖추고 있고 통계적 유의성을 보이는 검사-재검사 신뢰도를 보여 조사도구로서 유용한 것으로 판단된다. 요인분석의 결과 어느 요인에도 속하지 않는 문항이 있었고 이 문항의 수정 또는 누락이 필요한 것으로 생각되며, 영문판 EAT-26과 비교해볼 때, EAT-26KA의 요인구조는 상이하게 나타나 실제검사를 통해 얻어진 결과의 해석은 서구의 경우와 차이가 있을 것이 예상되었고 구조화된 면담을 통해 타당도의 확립이 필요한 것으로 생각되었다.