• 한국컴퓨터정보학회논문지
• /
• 제18권1호
• /
• pp.177-183
• /
• 2013

본 논문에서는 데이터 마이닝에서 분류 작업을 실시할 때 그 분류정확도을 측정하기 위해 Cohen의 kappa 계수와 weighted kappa 계수를 제안하였다. kappa 계수는 우연에 의해 생기는 분류를 보정하여 분류정확도을 측정하며 명목척도와 순서척도의 데이터에 대해 사용된다. 특히 순서척도의 데이터에서는 오분류의 크기를 가중치에 의해 정량화하여 분류정확도을 측정하는 weighted kappa 계수가 더 유용하게 사용된다. weighted kappa 계수 계산을 위해서는 2가지 가중치(일차형 가중치, 이차형 가중치)를 사용하였다.. 또한 실제 데이터인 지방간 데이터에 대해 kappa 계수와 weighted kappa 계수를 계산하여 비교하였다.

• East Asian mathematical journal
• /
• 제31권5호
• /
• pp.695-706
• /
• 2015

Let ${\mathcal{H}}$ be an infinite dimensional complex Hilbert space, and let $B({\mathcal{H}})$ be the algebra of all bounded linear operators on ${\mathcal{H}}$. Recall that an operator $T{\in}B({\mathcal{H})$ has property B(n) if ${\mid}T^n{\mid}{\geq}{\mid}T{\mid}^n$, $n{\geq}2$, which generalizes the class A-operator. We characterize the property B(n) of weighted shifts $S_{\lambda}$ over (${\eta},\;{\kappa}$)-type directed trees which appeared in the study of subnormality of weighted shifts over directed trees recently. In addition, we discuss the property B(n) of weighted shifts $S_{\lambda}$ over (2, 1)-type directed trees with nonzero weights are being distinct with respect to $n{\geq}2$. And we give some properties of weighted shifts $S_{\lambda}$ over (2, 1)-type directed trees with property B(2).

• 대한수학회보
• /
• 제40권2호
• /
• pp.341-349
• /
• 2003

Let {$x_{nk}\;$\mid$1\;\leq\;k\;\leq\;n,\;n\;\geq\;1$} be an array of random varianbles and $\{a_n$\mid$n\;\geq\;1\}\;and\;\{b_n$\mid$n\;\geq\;1} be a sequence of constants with $a_n\;>\;0,\;b_n\;>\;0,\;n\;\geq\;1. In this paper, for array of row negatively associated(NA) random variables, we establish a general weak law of large numbers (WLLA) of the form (${\sum_{\kappa=1}}^n\;a_{\kappa}X_{n\kappa}\;-\;\nu_{n\kappa})\;/b_n$ converges in probability to zero, as $n\;\rightarrow\;\infty$, where {$\nu_{n\kappa}$\mid$1\;\leq\;\kappa\;\leq\;n,\;n\;\geq\;1$} is a suitable array of constants.

• 대한수학회논문집
• /
• 제28권4호
• /
• pp.741-750
• /
• 2013

To study the operator inequalities, the notions of class A operators and quasi-class A operators are developed up to recently. In this paper, quasi-$A(n,{\kappa})$ class operator for $n{\geq}2$ and ${\kappa}{\geq}0$ is introduced as a new notion, which generalizes the quasi-class A operator. We obtain some structural properties of these operators. Also we characterize quasi-$A(n,{\kappa})$ classes for n and ${\kappa}$ via backward extension of weighted shift operators. Finally, we give a simple example of quasi-$A(n,{\kappa})$ operators with two variables.

• 응용통계연구
• /
• 제22권5호
• /
• pp.1073-1084
• /
• 2009

두 평정자가 R개의 순서형 반응 범주로 각 개체를 분류한 $R{\times}R$ 분할표에 대해, 불합치의 정도를 가중치로 부여한 가중 합치도 $H_{\omega}$를 제안하고, 최대 우도추정량 및 분산을 유도하였다. 또한 $2{\times}2$ 분할표에서 Feinstein과 Cicchetti(1990)가 제기한 마지막 역설을 새롭게 정의하고 증명하였으며, ${\kappa}$의 새로운 역설을 제기하고, ${\kappa}$와 주변분포의 전반적인 관계를 정리하였다.

• Journal of Nutrition and Health
• /
• 제29권9호
• /
• pp.1021-1027
• /
• 1996

This study examined the patterns of nutrient intakes measured by 1-, 3-, 7-day recalls and records as well as food frequency questionnaire among 59 females volunteers enrolled in the university in Seoul, Korea. Over a 4 month period, a modified Willett food frequency questionnaire was administered once, and a 24-hour dietary recall was conducted 12 times and a weighted dietary record 14 times. From these 12 recalls and 14 records, 1-, 3-, 7-day data were randomly selected. For energy and 11 nutrients, group mean intakes derived from food frequency questionnaire were higher than from recalls and records. Group mean intakes from recalls and records showed little differences depending on days of dietary studies and dietary methods. Measures of agreement were calculated by weighted kappa and intraclass correlation coefficient values calculated for quintile categories while comparing to the results of 26 days recalls and records. Weighted kappa values ranged from 0.11 for riboflavin to 0.36 for vitamin C for and 1-day recall, and from 0.21 for iron to 0.31 for energy for the 1-day record. Weighted kappa values were increased as the number of days of dietary studies increased (0.34-0.57 for the 3-day recalls, 0.27-0.50 for the 3-day records, 0.50-0.68 for the 7-day recalls, and 0.50-0.65 for the 7-day records). Weighted kappa values for food frequency questionnaire were higher than the 1-day data, but lower than the 3-and-7day data(0.34 for energy, 0.31 for iron and 0.22 for vitamin C). Intraclass correlation coefficients ranged from 0.21 for vitamin A to 0.57 for calcium. The degrees of agreement by different methods and days of dietary study are lower in our study compared to agreement by different methods and days of dietary study are lower in our study compared to those in previously published studies for Western populations, partly due to the differences of data analysis methods as well as of dietary patterns between those samples and ours.

• Journal of the Korean Data and Information Science Society
• /
• 제27권4호
• /
• pp.899-910
• /
• 2016

근사검정은 종종 표본이 작은 순서척도의 범주를 갖는 분할표를 분석할 때 그 p값이 과대추정 되거나 과소추정 되기 때문에 적절하지 못한 것으로 여겨진다. 본 논문에서는 순서화된 범주를 갖는 $k{\times}k$ 분할표에서 특정 범주에 대한 가중 일치도에 대해 정확한 p값과 재표본 기법에 의해 p값을 구하는 퍼뮤테이션 방법을 제시한다. 이를 위해 두 명의 평가자가 특정의 범주에서 얼마나 일치된 평가를 하는 지를 측정하기 위해 $Kv{\dot{a}}lseth$가 제안한 특정 범주에 대한 가중 일치도 (weighted specific-category kappa)를 사용한다. 사례 데이터로서 $3{\times}3$ 분할표 형태의 실제 데이터와 가상데이터 그리고 $4{\times}4$ 분할표 형태의 가상데이터를 이용하며, 정확한 퍼뮤테이션 p값과 재표본 퍼뮤테이션 p값 그리고 근사검정의 p값을 계산하여 비교한다.

• 한국추진공학회지
• /
• 제4권4호
• /
• pp.59-69
• /
• 2000

2차인 압축성/비압축성 Navier-Stokes 방정식은 이용하여 DCA 압푹기 익렬의 수치 해석을 수행하고 실험치와 비교 검토하였다. SIMPLE 알고리즘을 적용한 비압축성 코드는 대류항의 이산화에 하이브리드 도식을 진동해를 방지하기 위해 집중격자 기법을 사용하였다. 압축성 코드는 예조건화 기법 을 적용하였으며 공간 이산화출 위해 풍상 차분법을, 시간 적분을 위해서는 LU-SGS 기법을 사용하였다. 또한 난류 점성 유동장을 해석하기 위해 Baldwin-Lomax, standard $\kappa$ -$\varepsilon$, $\kappa$ -$\varepsilon$ Lam. Bremhorst, standard $\kappa$-$\omega$, $\kappa$ -$\omega$ SST 모델 등의 난류 모델을 적용하여 각 모델들의 특성을 살펴보았다.

• 대한수학회논문집
• /
• 제32권3호
• /
• pp.765-778
• /
• 2017

Finding an initial feasible solution on the central path is the main difficulty of feasible interior-point methods. Although, some algorithms have been suggested to remedy this difficulty, many practical implementations often do not use perfectly centered starting points. Therefore, it is worth to analyze the case that the starting point is not exactly on the central path. In this paper, we propose a weighted-path following interior-point algorithm for solving the Cartesian $P_{\ast}({\kappa})$-linear complementarity problems (LCPs) over symmetric cones. The convergence analysis of the algorithm is shown and it is proved that the algorithm terminates after at most $O\((1+4{\kappa}){\sqrt{r}}{\log}{\frac{x^0{\diamond}s^0}{\varepsilon}}\)$ iterations.

• 응용통계연구
• /
• 제25권5호
• /
• pp.719-731
• /
• 2012

Agreement analysis is conducted to assess reliability among rating results performed repeatedly on the same subjects by one or more raters. The kappa statistic is commonly used when rating scales are categorical. The simple and weighted kappa statistics are used to measure the degree of agreement between two raters, and the generalized kappa statistics to measure the degree of agreement among more than two raters. In this paper, we compare the performance of four different generalized kappa statistics proposed by Fleiss (1971), Conger (1980), Randolph (2005), and Gwet (2008a). We also examine how sensitive each of four generalized kappa statistics can be to the marginal probability distribution as to whether marginal balancedness and/or homogeneity hold or not. The performance of the four methods is compared in terms of the relative bias and coverage rate through simulation studies in various scenarios with different numbers of raters, subjects, and categories. A real data example is also presented to illustrate the four methods.