• Communications for Statistical Applications and Methods
• /
• v.19 no.2
• /
• pp.277-286
• /
• 2012

For credit assessment models, the ROC curves evaluate the classification performance using two univariate cumulative distribution functions of the false positive rate and true positive rate. In this paper, it is extended to two bivariate normal distribution functions of default and non-default borrowers; in addition, the bivariate ROC curves are proposed to represent the joint cumulative distribution functions by making use of the linear function that passes though the mean vectors of two score random variables. We explore the classification performance based on these ROC curves obtained from various bivariate normal distributions, and analyze with the corresponding AUROC. The optimal threshold could be derived from the bivariate ROC curve using many well known classification criteria and it is possible to establish an optimal cut-off criteria of bivariate mixture distribution functions.

• Journal of the Korean Society of Hazard Mitigation
• /
• v.8 no.6
• /
• pp.113-120
• /
• 2008

This study analyzed the spatial correlograms of minutely rainfall data with respect to various accumulation times. A bivariate mixed lognormal distribution was applied for rainfall modelling. A total of 26 minutely rainfall data sets from rain gauge stations in the central part of Korean peninsula were analyzed, also repeated for several storm types like Jang-Ma, typhoon and convective storms for their comparison. The accumulation times 1, 2, 3, 5, 10, 30 and 60 minutes were considered in this study. As results, it was found that the minutely rainfall data available was not good enough for estimating minutely rainfall intensity at ungaged locations. It seems more practical to use the hourly rainfall data with much higher rain gauge density, if proper methods for interpolation and data dis-aggregation are provided.

• Journal of the Korean Geotechnical Society
• /
• v.19 no.5
• /
• pp.145-154
• /
• 2003

In this study, the variation in excess pore pressure during dissipation is estimated by using successive cavity expansion theory and finite difference technique based on axisymmetric uncoupled linear consolidation theory with separate consideration of magnitude and initial distribution $\Delta{u}_{oct}$induced by changes of octahedral normal stress, and $\Delta{u}_{shear}$ induced by changes of octahedral shear stress. The coefficient of consolidation is also estimated by trial and error procedure until the predicted dissipation curve matches the measured curve at a typical degree of dissipation. The proposed method is applied to the results of miniature piezocone tests at Louisiana State University calibration chamber system. Based on the results of interpretation and the comparison with experimental measurements and those from other solutions, the prediction dissipation curves show a good match with those measured during dissipation tests and the values of coefficient of consolidation estimated by proposed method are more close to the range of laboratory measurements than those of other theories.

• Journal of Animal Science and Technology
• /
• v.55 no.5
• /
• pp.389-398
• /
• 2013

This study was conducted to investigate the genetic relationships among four reproductive traits. Data for this study were 7616 records from 1910 Landrace (L) and 10,454 records from 2283 Yorkshire (Y) in a closed nucleus swine herd. Traits considered on this study were gestation length (GL), total number of piglets born (TNB), wean to first service (WFS), and number of stillborn per litter (NSB). Heritabilities and genetic correlations were estimated by using the Bayesian inferences via Gibbs sampling in a four trait linear-threshold repeatability animal mixed model by designating NSB as a categorical trait in the L and Y purebred populations. Effects on the statistical model were considered for parity, contemporary group as fixed and service sire, permanent environmental, animal additive genetic effects as random. Estimates of heritability were 0.21, 0.23, 0.16, and 0.09 for GL, WFS, TNB, and NSB in the L population and 0.35, 0.16, 0.14 and 0.10 for corresponding traits in the Y population, respectively. Genetic correlation for GL was -0.59 and -0.28 with TNB and -0.58 and -0.17 with NSB in the L and Y populations, respectively. The NSB was positively correlated with TNB in the L and Y populations in genetic and environmental aspects. Therefore, the NSB should be taken into account in selecting sows for improving prolificacy of dam line breeding swine stock.