• Proceedings of the Korean Society for Information Management Conference
• /
• 2012.08a
• /
• pp.133-136
• /
• 2012

지적구조의 시각적인 표현을 위해서 오랫동안 다차원척도법이 사용되어 왔으나, 최근에는 패스파인더 네트워크와 같이 네트워크의 형태로 표현하는 방법이 확산되고 있다. 다차원척도법에 대해서는 생성된 2차원 지도의 품질을 평가하기 위한 합치도 지표로 STRESS나 RSQ를 비롯한 여러 종류가 사용되고 있는 반면, 네트워크 표현에 대해서는 마땅한 합치도 지표가 아직까지 제안되지 않았다. 이 글에서는 개체 간의 연관성을 네트워크로 표현할 경우에, 생성된 네트워크 구조를 평가하기 위한 합치도 지표로 NetRSQ를 제안하였다.

• Journal of the Korean Data and Information Science Society
• /
• v.14 no.4
• /
• pp.951-963
• /
• 2003

In this paper, we propose goodness of fit test statistics based on the estimated Kullback-Leibler information functions using the data from three step stress accelerated life test. This acceleration model is assumed to be a tampered random variable model. The power of the proposed test under various alternatives is compared with Kolmogorov-Smirnov statistic, Cramer-von Mises statistic and Anderson-Darling statistic.

• Journal of the Korean Data and Information Science Society
• /
• v.17 no.1
• /
• pp.41-52
• /
• 2006

Because most common assumption is normality in statistical analysis, testing normality is very important. The Q-Q plot is a powerful tool to test normality with full samples in statistical package. But the plot can't test normality in type-II censored samples. This paper proposed the modified the Q-Q plot and the modified normalized sample Lorenz curve(NSLC) for normality test in the type-II censored samples. Using the two Hodgkin's disease data sets and the type-II censored samples, we picture the modified Q-Q plot and the modified normalized sample Lorenz curve.

• Communications for Statistical Applications and Methods
• /
• v.19 no.6
• /
• pp.793-798
• /
• 2012

In this paper, we propose regression methods based on the likelihood function. We assume Arnold-Beaver Skew Normal(ABSN) errors in a simple linear regression model. It was shown that the novel method performs better with an asymmetric data set compared to the usual regression model with the Gaussian errors. The utility of a novel method is demonstrated through simulation and real data sets.

• Computational Structural Engineering : An International Journal
• /
• v.2 no.1
• /
• pp.1-10
• /
• 2002

Information on the distribution of the basic random variables is essential for the accurate evaluation of structural reliability. The usual method for determining the distribution is to fit a candidate distribution to the histogram of available statistical data of the variable and perform appropriate goodness-of-fit tests. Generally, such candidate distributions would have two parameters that may be evaluated from the mean value and standard deviation of the statistical data. In the present paper, a-parameter Gamma distribution, whose parameters can be directly defined in terms of the mean value, standard deviation and skewness of available data, is suggested. The flexibility and advantages of the distribution in fitting statistical data and its significance in structural reliability evaluation are identified and discussed. Numerical examples are presented to demonstrate these advantages.

• Asia-Pacific Journal of Business
• /
• v.9 no.2
• /
• pp.1-13
• /
• 2018

Physics simulation is an important part of many interactive 2D applications and collision detection and response is key component of this simulation. While methods for reducing the number of collision tests that need to be performed has been well researched, methods for performing the final checks with collision primitives have seen little recent development. This paper presents a new collision primitive, the n-arc, constructed from piecewise circular curves or biarcs. An algorithm for performing a collision check between these primitives is presented and compared to a convex polygon primitive. The n-arc is shown to exhibit similar, though slightly slower, performance to a polygon when no collision occurs, but is considerably faster when a collision does occur. The goodness of fit of the new primitive is also compared to a polygon. While the n-arc often gives a looser fit in terms of area, the continuous tangents of the n-arcs makes them a good choice for organic, soft or curved surfaces.

• Communications for Statistical Applications and Methods
• /
• v.9 no.1
• /
• pp.197-212
• /
• 2002

When assessing goodness-of-fit of a model, a small subset of deviating observations can give rise to a significant lack of fit. It is therefore important to identify such observations and to assess their effects on various aspects of analysis. A Cook's distance measure is usually used to detect influential observation. But it sometimes is not fully effective in identifying truly influential set of observations because there may exist masking or swamping effects. In this paper we confine our attention to influential subset In GLMs such as logistic regression models and loglinear models. We modify a backwards stepping algorithm, which was originally suggested for detecting outlying cells in contingency tables, to detect influential observations in GLMs. The algorithm consists of two steps, the identification step and the testing step. In identification step we Identify influential observations based on influencial measures such as Cook's distances. On the other hand in testing step we test the subset of identified observations to be significant or not Finally we explain the proposed method through two types of dataset related to logistic regression model and loglinear model, respectively.

• Structural Engineering and Mechanics
• /
• v.28 no.3
• /
• pp.263-280
• /
• 2008

Information on the distribution of the basic random variable is essential for the accurate analysis of structural reliability. The usual method for determining the distributions is to fit a candidate distribution to the histogram of available statistical data of the variable and perform approximate goodness-of-fit tests. Generally, such candidate distribution would have parameters that may be evaluated from the statistical moments of the statistical data. In the present paper, a cubic normal distribution, whose parameters are determined using the first four moments of available sample data, is investigated. A parameter table based on the first four moments, which simplifies parameter estimation, is given. The simplicity, generality, flexibility and advantages of this distribution in statistical data analysis and its significance in structural reliability evaluation are discussed. Numerical examples are presented to demonstrate these advantages.

• Journal of the Korean Society of Manufacturing Technology Engineers
• /
• v.18 no.5
• /
• pp.536-543
• /
• 2009

In this paper the mean stress effects on the probability distribution of the random grown crack size at a specified loading cycle are studied through the fatigue crack propagation tests, which are conducted on the specimens of magnesium alloy under four different stress ratios. Through 80 replicates the probability distributions of the grown crack size are obtained. The goodness-of-fit for probability distributions of the random grown crack size are investigated by Anderson-Darling test and the best fit for those probability distributions is found to be a 3-parameter Weibull distribution. The effects of the mean stress on the probability distribution of the random grown crack size are also estimated.

• Journal of Korean Society for Quality Management
• /
• v.46 no.1
• /
• pp.95-112
• /
• 2018

Purpose: To find out the appropriate probability distribution of credit card usage behavior by considering the relationship among income, expenditure and credit card usage amount. Such relationship is enabled by Korea's especially high penetration of credit card. Method: Goodness-of-fit test and effect size statistic W were used to identify the distribution of income and credit card usage amount. A simulation model is introduced to generate the credit card transactions on individual user level. Result: The three data sets for testing had either passed the chi-square test or showed low W values, meaning they follow the exponential distribution. And the exponential distribution turned out to fit the data sets well. The r values were very high. Conclusion: The credit card usage behavior, denoted as the counts of users by usage amount band, follows the exponential distribution. This distribution is easy to manipulate, has a variety of applications and generates important business implications.