• Nuclear Engineering and Technology
• /
• 제33권3호
• /
• pp.270-282
• /
• 2001

A simple measure of uncertainty importance based on normalized metric distance to quantify the entire change of cumulative distribution functions (CDFs) has been developed for use in probability safety assessments (PSAs). The metric distance measure developed in this study reflects the relative impact of distributional changes of inputs on the change of an output distribution, white most of the existing uncertainty importance measures reflect the magnitude of relative contribution of input uncertainties to the output uncertainty. Normalization is made to make the metric distance measure a dimensionless quantity. The present measure has been evaluated analytically for various analytical distributions to examine its characteristics. To illustrate the applicability and strength of the present measure, two examples are provided. The first example is an application of the present measure to a typical problem of a system fault tree analysis and the second one is for a hypothetical non-linear model. Comparisons of the present result with those obtained by existing uncertainty importance measures show that the metric distance measure is a useful tool to express the measure of uncertainty importance in terms of the relative impact of distributional changes of inputs on the change of an output distribution.

• 한국원자력학회:학술대회논문집
• /
• 한국원자력학회 1996년도 추계학술발표회논문집(1)
• /
• pp.415-420
• /
• 1996

The main objective of the present study is to propose a new measure of uncertainty importance based on distributional sensitivity analysis. The new measure is developed to utilize a metric distance obtained from cumulative distribution functions (cdfs). The measure is evaluated for two cases: one is a cdf given by a known analytical distribution and the other given by an empirical distribution generated by a crude Monte Carlo simulation. To study its applicability, the present measure has been applied to two different cases. The results are compared with those of existing three methods. The present approach is a useful measure of uncertainty importance which is based on cdfs. This method is simple and easy to calculate uncertainty importance without any complex process. On the basis of the results obtained in the present work, the present method is recommended to be used as a tool for the analysis of uncertainty importance.

• 대한산업공학회지
• /
• 제13권2호
• /
• pp.79-85
• /
• 1987

A unified reliability importance of the component in the network is considered. A measure is presented by considering the traffic between vertex pairs. An algrithm for this measure is also presented by using the factoring algorithm.

• 패션비즈니스
• /
• 제19권6호
• /
• pp.127-138
• /
• 2015

Due to the virtual nature of online businesses, online apparel stores need to enhance the consumer experience by utilizing store design attributes to provide their customers relevant and sufficient information. Since online apparel stores mainly communicate with their customers virtually and digitally, it is important to understand how consumers perceive and react to different design attributes commonly found on apparel stores' online websites. The purpose of this paper is to examine the importance of design attributes commonly found on online apparel stores' websites using three different importance measurements. The design attributes examined in this study include enlarged pictures, product detail pictures, product reviews by other buyers, coordinating items, and size measurement charts. The three different measurements used in this study include two direct measures and one indirect measure using conjoint analysis. Across the three different measures, both the men and women indicated that enlarged pictures represent the most important design attribute when they purchase clothes online followed by size measurement charts and they considered the availability of coordinating items the least important design attribute.

• 한국경영과학회:학술대회논문집
• /
• 대한산업공학회/한국경영과학회 1990년도 춘계공동학술대회논문집; 한국과학기술원; 28 Apr. 1990
• /
• pp.272-288
• /
• 1990

The uncertainty quantification process in probabilistic Risk Assessment usually involves a specification of the uncertainty in the input data and the propagation of this uncertainty to the final risk results. The distributional sensitivity analysis is to study the impact of the various assumptions made during the quantification of input parameter uncertainties on the final output uncertainty. The uncertainty importance of input parameters, in this case, should reflect the degree of changes in the whole output distribution and not just in a point estimate value. A measure of the uncertainty importance is proposed in the present paper. The measure is called the distributional sensitivity measure(DSM) and explicitly derived from the definition of the Kullback's discrimination information. The DSM is applied to three typical discrimination information. The DSM is applied to three typical cases of input distributional changes: 1) Uncertainty is completely eliminated, 2) Uncertainty range is increased by a factor of 10, and 3) Type of distribution is changed. For all three cases of application, the DSM-based importance ranking agrees very well with the observed changes of output distribution while other statistical parameters are shown to be insensitive.

• 한국원자력학회:학술대회논문집
• /
• 한국원자력학회 2004년도 요약집 춘계학술발표대회
• /
• pp.230-230
• /
• 2004

• 한국정보시스템학회지:정보시스템연구
• /
• 제17권3호
• /
• pp.25-37
• /
• 2008

In a fault tree analysis, an uncertainty importance measure is often used to assess how much uncertainty of the top event probability (Q) is attributable to the uncertainty of a basic event probability ($q_i$), and thus, to identify those basic events whose uncertainties need to be reduced to effectively reduce the uncertainty of Q. For evaluating the measures suggested by many authors which assess a percentage change in the variance V of Q with respect to unit percentage change in the variance $v_i$ of $q_i$, V and ${\partial}V/{\partial}v_i$ need to be estimated analytically or by Monte Carlo simulation. However, it is very complicated to analytically compute V and ${\partial}V/{\partial}v_i$ for large-sized fault trees, and difficult to estimate them in a robust manner by Monte Carlo simulation. In this paper, we propose a method for evaluating the measure using discretization technique and Monte Carlo simulation. The proposed method provides a stable uncertainty importance of each basic event.

• Earthquakes and Structures
• /
• 제22권2호
• /
• pp.157-168
• /
• 2022

A moment-independent importance measure analysis approach was introduced to quantify the effects of structural uncertainty parameters on probabilistic seismic demands of simply supported girder bridges. Based on the probability distributions of main uncertainty parameters in bridges, conditional and unconditional bridge samples were constructed with Monte-Carlo sampling and analyzed in the OpenSees platform with a series of real seismic ground motion records. Conditional and unconditional probability density functions were developed using kernel density estimation with the results of nonlinear time history analysis of the bridge samples. Moment-independent importance measures of these uncertainty parameters were derived by numerical integrations with the conditional and unconditional probability density functions, and the uncertainty parameters were ranked in descending order of their importance. Different from Tornado diagram approach, the impacts of uncertainty parameters on the whole probability distributions of bridge seismic demands and the interactions of uncertainty parameters were considered simultaneously in the importance measure analysis approach. Results show that the interaction of uncertainty parameters had significant impacts on the seismic demand of components, and in some cases, it changed the most significant parameters for piers, bearings and abutments.

• 품질경영학회지
• /
• 제37권1호
• /
• pp.56-68
• /
• 2009

A measure for service quality of university dormitory is developed to gain a better understanding of the quality issues that impact on students' experiences and to improve service quality. Literature survey, thorough discussion with staff members and a pilot test are utilized to elicit attributes of service quality. Factor analysis is used to group the service quality attributes into dimensions. The resulting measure is consisted of 6 dimensions; competence, attitude, facility, amenity, security and discipline. Importance-performance analysis is utilized to verify which factors to be focused on with high Priority to improve dormitory service.

• 정보처리학회논문지:소프트웨어 및 데이터공학
• /
• 제3권1호
• /
• pp.37-42
• /
• 2014

연관 규칙들을 이용한 분류학습은 최근 활발히 연구되는 분야의 하나이다. 이러한 연관 규칙을 이용한 분류에는 연관 규칙들에 대한 수치적 중요도를 계산하는 것이 중요하다. 본 논문에서는 정보 이론을 사용한 H measure 라는 새로운 규칙 중요도 기법을 제안한다. 구체적으로 Hellinger 변량을 이용하여 연관규칙의 중요도를 계산한다. 제안된 H measure 의 다양한 특성들을 분석하였으며 또한 이러한 H measure를 이용한 분류학습의 성능을 다른 규칙 measure를 이용한 분류학습의 성능과 비교하였다.