• Proceedings of the Korea Water Resources Association Conference
• /
• 2010.05a
• /
• pp.571-575
• /
• 2010

Drought is a natural hazard with different properties that are usually dependent to each other. Therefore, a multivariate model is often used for drought frequency analysis. The Copula based bivariate drought severity and duration frequency analysis is applied in the current study in order to show the effect of tail behavior of drought severity and duration on the selection of a copula function for drought bivariate frequency analysis. Four copula functions, namely Clayton, Gumbel, Frank and Gaussian, were fitted to drought data of four stations in Iran and Canada in different climate regions. The drought data are calculated based on standardized precipitation index time series. The performance of different copula functions is evaluated by estimating drought bivariate return periods in two cases, [$D{\geq}d$ and $S{\geq}s$] and [$D{\geq}d$ or $S{\geq}s$]. The bivariate return period analysis indicates the behavior of the tail of the copula functions on the selection of the best bivariate model for drought analysis.

• Journal of the Korean Data and Information Science Society
• /
• v.25 no.1
• /
• pp.219-226
• /
• 2014

In this paper, we derived estimators of reliability P(Y < X) and the distribution of ratio in the bivariate exponential density. We also considered the means and variances of M = max{X,Y} and m = min{X,Y}. We finally presented how E(M), E(m), Var(M) and Var(m) are varied with respect to the ones in the bivariate exponential density.

• Journal of the Korean Data and Information Science Society
• /
• v.21 no.1
• /
• pp.173-177
• /
• 2010

It is well known that if the parent distribution has a nonnegative support and has increasing failure rate, then all the order statistics have increasing failure rate (IFR). The result is not necessarily true in the case of bivariate distributions with dependent structures. In this paper we consider a symmetric bivariate exponential distribution and show that, two marginal distributions are IFR and the distributions of the minimum and maximum are constant failure rate and IFR, respectively.

• Journal of the Korean Statistical Society
• /
• v.25 no.1
• /
• pp.133-144
• /
• 1996

We, in this paper, propose a nonparametric estimator of bivariate mean residual life function based on Lin and Ying's (1993) bivariate survival function estimator of paired failure times under univariate censoring and prove the uniform consistency and the weak convergence result of this estimator. Through Monte Carlo simulation, the performances of the proposed estimator are tabulated and are illustrated with the skin grafts data.

• Journal of the Korean Data and Information Science Society
• /
• v.23 no.6
• /
• pp.1213-1222
• /
• 2012

We shall derive the moment of the ratio Y/(X + Y) and the reliability P(X < Y ), and then observe the skewness of the ratio in a bivariate Pareto density function of (X, Y). And we shall consider an approximate MLE of parameters in the bivariate Pareto density function.

• Honam Mathematical Journal
• /
• v.35 no.1
• /
• pp.29-35
• /
• 2013

The Fisher information matrix plays a significant role i statistical inference in connection with estimation and properties of variance of estimators. Using Bivariate Lomax distribution, we can define "statistical model" and drive the Fisher information matrix of Bivariate Lomax distribution. In this paper, we correct the wrong of the paper [7].

• Journal of applied mathematics & informatics
• /
• v.24 no.1_2
• /
• pp.221-230
• /
• 2007

The exponential and the gamma distributions have been the traditional models for drought duration and drought intensity data, respectively. However, it is often assumed that the drought duration and drought intensity are independent, which is not true in practice. In this paper, an application of the bivariate gamma exponential distribution is provided to drought data from Nebraska. The exact distributions of R=X+Y, P=XY and W=X/(X+Y) and the corresponding moment properties are derived when X and Y follow this bivariate distribution.

• Journal of the Korean Data and Information Science Society
• /
• v.24 no.4
• /
• pp.895-900
• /
• 2013

In this research, we propose a simple bivariate zero inflated negative binomial regression model with different dispersion for bivariate count data with excess zeros. An application to the demand for health services shows that the proposed model is better than existing models in terms of log-likelihood and AIC.

• Journal of the Korean Statistical Society
• /
• v.5 no.1
• /
• pp.3-18
• /
• 1976

To test simultaneous significance of more than two linear regression coefficients, we can consider multivariate-t tests with critical regions in t-space instead of F-tests where t-values are t-statistics of significance tests of one coefficient. In this paper bivariate-t distributions and bivariate-t tests of two coefficients such as maxmod, minmod, one-tailed maxmod and one-tailed minmod tests are studied. Through the calculation of powers of test, it is learned that in some cases bivariate-t test are more powerful than F-tests.

• Journal of the Korean Statistical Society
• /
• v.36 no.3
• /
• pp.349-355
• /
• 2007

In this paper we study some important properties of the bivariate generalized hypergeometric probability (BGHP) distribution by establishing the existence of all the moments of the distribution and by deriving recurrence relations for raw moments. It is shown that certain mixtures of BGHP distributions are again BGHP distributions and a limiting case of the distribution is considered.