• Communications for Statistical Applications and Methods
• /
• v.1 no.1
• /
• pp.206-214
• /
• 1994

본 논문은 1994학년도 부터 부활된 대학별고사 실시에 따른 문항분석, 표준 점수제 그리고 검사동등화의 문제점을 지적하기 위하여 교육측정이론의 기본 개념을 소개하는데 있다. 대학별고사의 타당성과 신뢰성을 보장받기 위하여는 양질의 문항제작이 우선하여야하며, 이를 위하여 문항분석은 종전에 사용하던 고전검사이론 보다는 문항반응이론을 이용하는 것이 바람직하다. 문항반응이론에 의한 문항분석은 피험자 집단의 특성에 의하여 문항특성이 달리 분석되지 않는 특징을 지니고 있기 때문이다. 문항이 논술형일 경우 채점자간 신뢰도와 채점자 내 신뢰도를 간과하여서는 안될 것이다. 다양한 선택과목을 채택하는 대학별 고사에서 입학 사정을 위하여 원점수를 사용하거나, 표준점수 혹은 검사동등화 방법을 이용하고 있으나 이는 교육측정이론에 위배된다. 다른 과목에 대한 인가의 능력을 상대비교 할 수 없으며, 표준점수와 검사동등화는 동일 능력에 대한 상대비교를 위한 방법이다. 특히 검사동등화는 동일 특성, 공정성, 모교집단 불변성, 대칭성을 전제한다. 표준점수제에 의하여 수험생들의 다른 능력을 상대 비교하는 방법은 다른 능력이 점수로 표현되기 때문에 가능하나 그 점수가 무엇을 의미하는 가를 분석할 때는 교육평가의 기본 철학에도 위배된다.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2002.05a
• /
• pp.77-80
• /
• 2002

이상치를 포함한 학습 데이터의 군집화 전략은 일반적으로 이상치를 포함하여 학습하거나, 이상치를 제거하는 두 가지 선택이 가능하다. 이상치를 제거하지 않고 학습에 반영시켜야 할 경우 한 개 또는 소수의 이상치가 독자적인 군집을 형성하거나 객관적인 군집화를 방해하는 문제가 발생할 수 있다. 이 때 주어진 학습 데이터의 군집 결과가 이상치의 영향으로부터 벗어나기 위해 원래의 학습 데이터에 대한 변환 작업을 거친 후 군집화를 수행할 수 있다. 이러한 변환 방법으로서 본 논문에서는 차원 축소의 기법으로 알려진 인자 분석의 점수를 사용하였다. 인자 점수로 변환된 학습 데이터에 대해 계층적 군집화, K-means 그리고 자기조직화 지도 등과 같은 군집화 알고리즘을 적용하면 이상치가 자신만의 군집을 별도로 형성하지 않고 다른 학습 데이터의 군집에 소속되면서 이상회의 영향으로부터 벗어남을 실험을 통하여 확인하였다.

• Journal of KIISE:Computer Systems and Theory
• /
• v.34 no.5_6
• /
• pp.187-194
• /
• 2007

A local alignment algorithm does comparing two strings and finding a substring pair with size l and similarity s. To find a pair with both sufficient size and high similarity, existing normalization approaches maximize the ratio of the similarity to the size. In this paper, we introduce normalization by functions that maximizes f(s)/g(l), where f and g are non-decreasing functions. These functions, f and g, are determined by experiments comparing DNA sequences. In the experiments, our normalization by functions finds appropriate local alignments. For the previous algorithm, which evaluates the similarity by using the longest common subsequence, we show that the algorithm can also maximize the score normalized by functions, f(s)/g(l) without loss of time.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.18 no.7
• /
• pp.275-283
• /
• 2017

The purpose of this study is to investigate the evaluation scores among various evaluators such as tutor, peer, and self evaluation, the lenience and central tendency among these various evaluators, and the correlation between evaluation scores among the various evaluators and the paper test score in the Problem-Based Learning (PBL). For the research, tutor, peer, and self evaluations were carried out with PBL effectiveness evaluation tool for the 105 nursing students taking classes using PBL, and paper test was conducted after the PBL class ended. The collected data were analyzed by SPSS WIN (V. 12.0). The study results showed that peer and self evaluation scores were higher than tutor evaluation score. Also lenience and central tendency were higher and stronger in peer and self evaluation than in tutor evaluation. Paper test score had a significant correlation with tutor evaluation score, but not with peer or self evaluation scores.

• 한국데이터정보과학회:학술대회논문집
• /
• 2006.04a
• /
• pp.25-32
• /
• 2006

본 논문에서는 민감한 사항 특히 청렴도 조사에서 금품 향응제공여부 등과 같이 응답자들이 정직하게 응답하지 않을 수 있는 질문들에 대하여 응답자의 비밀을 노출시키지 않고서 양질의 보다 정확한 정보를 얻을 수 있는 확률화응답기법을 적용하여 실제 청렴도 점수를 추정하고자 한다. 본 연구는 청렴도 조사와 같은 민감한 질문에 대하여 응답자에게 심리적 안정감을 주어 거짓응답을 줄이면서 보다 진실에 가까운 청렴도 점수를 구할 수 있는 사례를 제시한다.

• The Korean Journal of Applied Statistics
• /
• v.10 no.2
• /
• pp.275-291
• /
• 1997

In the case that an abstract concept was measured indirectly by using its indicators, many researcher have obtained its score by using the simple mean, the first principal component, or the first factor, etc. In this paper, an scoring method named as the representative component scoring system was suggested as an alternative and its validity and applicability were studied.

• Journal of the Korean Society of Radiology
• /
• v.10 no.4
• /
• pp.285-290
• /
• 2016

The purpose of our study was to retrospectively evaluate the cause of a decreased calcium score of follow-up studies on coronary artery calcium scores (CACs) computed tomography (CT). The subjects were healthy 100 people(85 males $60.6{\pm}6.9$ years, 15 females $67.2{\pm}7.3$ years). The subjects decreased CACs were divided into 4 subgroups depending on Agatston classification, minimal (1-10), mild (11-100), moderate (101-400), severe (400<). As a result of decreased CACs were scan location disagreement 51%, motion artifact 26%, equipment changes 14%, operator mistakes 5%, input miss 2%, image loss 1%, arrhythmia 1%. In the mild group, the most common decreased CACs were 49 people. In the minimal group, the most significant variation reduction has occurred to 6 people. Scan location disagreement was considered a partial volume effects due to the scan starting position. It showed less than 100 CACs a high variation (19.7%) in more than 100 CACs, a lower variation (2.2%), these could be seen that the variation range is different that can be tolerated according to the calcification score. Motion artifact factor was found in 26%, which is so closely related to the preceding tests that affect the higher heart rate like this pulmonary function test, exercise stress test.

• Korean Journal of Construction Engineering and Management
• /
• v.16 no.4
• /
• pp.98-106
• /
• 2015

VE aims at reduction in a budget, improvement of function, structural safety and quality security for public construction projects. However, there is possibility for the structural safety and quality security review to be insufficient because related regulations are mostly composed of analysis on economic efficiency of design. In addition, due to the misconception about VE as a cost saving methodology, an alternative is being presented which still focuses mainly on cost saving, but with no objective evaluation of function related to cost. In order to improve this, the government adopted the reduction of life cycle cost and proposal of value improvement, and let people specify the cost and function of the original plan versus the alternative plan, and the value changes between them. However, it is written mainly into practical convenience rather than theoretical basis since a specific way is not suggested. The current method sets a different starting point by applying the attributional difference of function and cost. Furthermore, an evaluation standard for correlating is an important element in rational decision making for assessing and choosing an alternative. This paper analyzes the process and method of function & cost scoring when performing VE and suggests a mathematical normalization model in order to support rational decision making when selecting an optimum plan.

• Proceedings of the Korean Institute of Navigation and Port Research Conference
• /
• 2011.06a
• /
• pp.147-149
• /
• 2011

본 연구는 선박사고의 발생과 선박이 지닌 특성(선종/톤수/국적)간에 연관이 있을 것으로 가정하여 관제전문가를 대상으로 하는 설문조사 및 사고통계 분석을 통해 각 특성 별 가중치를 계량화하여 기준점수표를 마련하고자 하였다. 가중치 산정은 분석적 계층화 과정(AHP: Anlaytic Hierachy Process)를 거쳤으며 이를 통해 얻은 기준점수표를 바탕으로 인천항의 1일간 입항선에 적용하여 결과를 준사고이력 보유선박군의 점수와 비교하여 양 집단 간 높은 위험지수 해당선박 존재 비율을 비교하였다.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.22 no.2
• /
• pp.205-211
• /
• 2012

The retrieved documents have to be transformed into proper data structure for the clustering algorithms of statistics and machine learning. A popular data structure for document clustering is document-term matrix. This matrix has the occurred frequency value of a term in each document. There is a sparsity problem in this matrix because most frequencies of the matrix are 0 values. This problem affects the clustering performance. The sparseness of document-term matrix decreases the performance of clustering result. So, this research uses the factor score by factor analysis to solve the sparsity problem in document clustering. The document-term matrix is transformed to document-factor score matrix using factor scores in this paper. Also, the document-factor score matrix is used as input data for document clustering. To compare the clustering performances between document-term matrix and document-factor score matrix, this research applies two typed matrices to self organizing map (SOM) clustering.