• 제목/요약/키워드: Criterion Factor

검색결과 764건 처리시간 0.025초

A Bayes Rule for Determining the Number of Common Factors in Oblique Factor Model

  • Kim, Hea-Jung
    • Journal of the Korean Statistical Society
    • /
    • 제29권1호
    • /
    • pp.95-108
    • /
    • 2000
  • Consider the oblique factor model X=Af+$\varepsilon$, with defining relation $\Sigma$$\Phi$Λ'+Ψ. This paper is concerned with suggesting an optimal Bayes criterion for determining the number of factors in the model, i.e. dimension of the vector f. The use of marginal likelihood as a method for calculating posterior probability of each model with given dimension is developed under a generalized conjugate prior. Then based on an appropriate loss function, a Bayes rule is developed by use of the posterior probabilities. It is shown that the approach is straightforward to specify distributionally and to imploement computationally, with output readily adopted for constructing required cirterion.

  • PDF

Bayesian Testing for Independence in Bivariate Exponential Model

  • Cho, Jang-Sik
    • Journal of the Korean Data and Information Science Society
    • /
    • 제17권2호
    • /
    • pp.521-527
    • /
    • 2006
  • In this paper, we consider the Bayesian hypotheses testing for independence in bivariate exponential model. In Bayesian testing problem, we use the noninformative priors for parameters which are improper and are defined only up to arbitrary constants. And we use the recently proposed hypotheses testing criterion called the fractional Bayes factor. Also we give some numerical results to illustrate our results.

  • PDF

Bayesian Approach for Independence Test in Bivariate Exponential Model

  • 조장식
    • 한국데이터정보과학회:학술대회논문집
    • /
    • 한국데이터정보과학회 2006년도 PROCEEDINGS OF JOINT CONFERENCEOF KDISS AND KDAS
    • /
    • pp.327-333
    • /
    • 2006
  • In this paper, we consider the Bayesian hypotheses testing for independence in bivariate exponential model. In Bayesian testing problem, we use the noninformative priors for parameters which are improper and are defined only up to arbitrary constants. And we use the recently proposed hypotheses testing criterion called the fractional Bayes factor. Also we give some numerical results to illustrate our results.

  • PDF

항공위협의 역사적 고찰과 항공보안 관리 연구 (Historical Review on Aviation Threats and Management of Aviation Security)

  • 안재형;강자영
    • 한국항공운항학회지
    • /
    • 제12권2호
    • /
    • pp.29-42
    • /
    • 2004
  • This paper performs a historical review on aviation threats through a statistical survey. The impact factor, as a criterion to measure the seriousness of the threats occurred, is introduced and evaluated based on annual totals of hijacking and fatalities against aircraft 1948 through 2004. Also, the paper suggests processes of evaluating the level of threats and key functions for an effective management of the aviation security.

  • PDF

Optimal Run Orders in Factorial Designs

  • Park, Sung H.;Lee, Jae W.
    • Journal of the Korean Statistical Society
    • /
    • 제15권1호
    • /
    • pp.31-45
    • /
    • 1986
  • It is often necessary to obtain some run orders in factorial designs which have a small number of factor level changes and a small linear time trend. In this paper we propose an algorithm to find optimal or near-optimal run orders for $2^4, 2^5, 3^2$ and $2\cdot 3^2$ factorial designs under the criterion that the number of factor level changes and the linear time trend should be simultaneously small.

  • PDF

테러인식 척도의 확인적요인분석 (A Confirmatory Factor Analysis on the Awareness of Terrorism Scale in Korea)

  • 정종운;곽대경;심혜인
    • 시큐리티연구
    • /
    • 제61호
    • /
    • pp.235-254
    • /
    • 2019
  • 이 연구는 테러 인식에 대한 척도에 대하여 확인적 요인분석(Confirmatory Factor Analysis)을 사용하여 요인구조를 확인함으로써, 한국형 테러인식을 측정하는데 적합한 척도인지에 대한 정보를 제공하고자 하였다. 최오호 외(2016) 「국회청사 보안강화 방안에 관한 연구」에서 수집된 국회공무원을 대상으로 한 176명의 자료를 분석에 활용하였다. 연구결과 테러인식의 총 14문항으로 구성된 척도에서 테러우려 3문항, 테러발생인식 5문항, 테러대응 3문항, 테러대비장치 3문항의 요인별 모형적합도의 적합도 지수는 기준치를 충족하여 적합한 것으로 나타났다. 표준화회귀계수도 기준 값을 상회하고 있어 적합한 것으로 나타났다. 잠재변수 테러우려가 평균분산추출값의 기준 값에 다소 미치지 못한 것을 확인하였으나, 그 이외의 값은 적합한 것으로 나타나 집중타당도가 확보된 것으로 보았다. 이를 보충하고자 판별타당도를 검증한 결과 척도 수준이 적합한 것으로 나타났다. 따라서 이 연구에서 검증한 테러인식문항을 포함하여 테러와 관련된 이론과 선행연구를 바탕으로 테러인식을 측정하기 위한 적합한 문항으로 구성된 설문지가 개발되어 측정될 필요가 있다.

불평형 전원전압을 갖는 정류시스템에서 출력필터에 따른 입력 특성 분석 (Analysis of Input Characteristic in the Rectifier for Output Filter with Unbalanced Supply Voltages)

  • 강수현;김상훈
    • 산업기술연구
    • /
    • 제25권B호
    • /
    • pp.195-202
    • /
    • 2005
  • The rectifier characteristics and the quality of the input current worsens with the increase of unbalances or harmonics of the supply voltages. Rectifier input current harmonics interfere with proper power system operation, reduce rectifier power factor, and limit the power available from a given source. It is of importance to select appropriately the rectifier's output filter inductance to determine the rectifier input current waveform, the input current harmonics, and the power factor. This paper presents a quantitative analysis of single and three phase rectifier input current harmonics, total harmonic distortion, and power factor as a function of the output filter inductance under balanced and unbalanced conditions. Also, its performance under the supply voltage including harmonics be investigated. These results provide a reference for selecting reasonable rectifier's output filter inductance for given harmonics or power factor criterion.

  • PDF

류마티스 관절염환자용 다차원적 피로척도의 타당도 및 신뢰도 (Validity and Reliability of Translated Multidimensional Assessment of Fatigue Scale for the Patients with Rheumatoid Arthritis)

  • 이경숙;이은옥
    • 근관절건강학회지
    • /
    • 제5권2호
    • /
    • pp.206-221
    • /
    • 1998
  • The purpose of this study was to validate translated Multidimensional Assessment of Fatigue(MAF) scale. The scale is a 16-item scale that measures four dimensions of fatigue : severity, distress, impact, timing. Fourteen items are numerical rating scales and 2 items have multiple choice responses. Data were collected from the 137 patients with rheumatoid arthritis after content validation. Criterion validity was tested by correlation coefficient with Piper Fatigue Scale, which resulted in 0.7573(p<.0000). Construct validity was tested by item analysis and factor analysis. Corrected item-total correlation coefficients were 0.63-0.88. And factor analysis showed 2 factors : fatigue degree factor and fatigue impact factor. These two factors explained 73.5% of total variance. Reliability of internal consistency was 0.96 in Cronbach's alpha. Further validation study is necessary in each factor in other settings with other subjects.

  • PDF

A STATISTICAL APPROACH FOR DERIVING KEY NFC EVALUATION CRITERIA

  • Kim, S.K.;Kang, G.B.;Ko, W.I.;Youn, S.R.;Gao, R.X.
    • Nuclear Engineering and Technology
    • /
    • 제46권1호
    • /
    • pp.81-92
    • /
    • 2014
  • This study suggests 5 evaluation criteria (safety and technology, environmental impact, economic feasibility, social factors, and institutional factors) and 24 evaluation indicators for a NFC (nuclear fuel cycle) derived using factor analysis. To do so, a survey using 1 on 1 interview was given to nuclear energy experts and local residents who live near nuclear power plants. In addition, by conducting a factor analysis, homogeneous evaluation indicators were grouped with the same evaluation criteria, and unnecessary evaluation criteria and evaluation indicators were dropped out. As a result of analyzing the weight of evaluation criteria with the sample of nuclear power experts and the general public, both sides recognized safety as the most important evaluation criterion, and the social factors such as public acceptance appeared to be ranked as more important evaluation criteria by the nuclear energy experts than the general public.

해상교통환경 위험요소 평가기준 설정을 위한 기초 연구 (A Basic Study on Assessment Criterion of the Risk Factor for the Marine Traffic Environment)

  • 김철승;이홍훈
    • 해양환경안전학회지
    • /
    • 제18권5호
    • /
    • pp.431-438
    • /
    • 2012
  • 해상교통환경의 위험도를 평가하기 위한 기술로는 대표적으로 FSA, PAWSA, IWRAP 등이 있으며, 이러한 기술의 개발을 위해서는 해상교통환경에 적합한 위험요소를 선정하고 이에 대한 평가기준이 마련되어야 한다. 기존 기술에서 위험도는 사고의 출현빈도와 이로 인한 영향의 곱으로 정의되어 이에 따라 사고의 출현빈도 및 영향에 해당되는 위험요소들이 각각 구분되어 선정되고 있었다. 그러나, 본 연구는 각 위험요소에 요소별 출현빈도와 영향을 포함하여 이들의 합으로 위험도를 정의함으로써 기존 기술에서 사고의 영향에 해당되었던 위험요소들을 제외하는 한편, 위험요소 분류체계에 관한 기존 연구의 사례 검토를 통하여 위험도를 구성하는 위험요소를 20가지로 추출한 후 유사한 성격에 따라 5가지 카테고리로 분류하였다. 또한, 선정된 각 위험요소에 대하여 관련 통계자료 등을 이용 실용적으로 용이하게 평가할 수 있는 기준을 제시하여, 향후 국내 해상교통환경에 적합한 위험도 평가모델의 개발을 위한 기초를 마련하였다.