• Communications for Statistical Applications and Methods
• /
• 제9권3호
• /
• pp.799-808
• /
• 2002

This paper aims at selecting the multi-response design with γ responses minimizing the bias error caused by fitting inadequate models to responses, where the first order models are fitted to Ρ responses fearing the quadratic bias, while to other γ- Ρ responses, the quadratic models are fitted fearing the cubic biases in the cuboidal region of interest. Under the assumption of symmetric design, by minimizing the criterion which represents the amount of error caused by fitting inadequate models, the optimum design was found to be the one having the design moments of second order and the fourth order as 1/3 and l/5, respectively. Examples of the design meeting the required conditions are given for illustration.

• Communications for Statistical Applications and Methods
• /
• 제21권3호
• /
• pp.235-243
• /
• 2014

Constrained Bayesian estimates overcome the over shrinkness toward the mean which usual Bayes and empirical Bayes estimates produce by matching first and second empirical moments; subsequently, a constrained Bayes estimate is recommended to use in case the research objective is to produce a histogram of the estimates considering the location and dispersion. The well-known squared error loss function exclusively emphasizes the precision of estimation and may lead to biased estimators. Thus, the balanced loss function is suggested to reflect both goodness of fit and precision of estimation. In insurance pricing, the accurate location estimates of risk and also dispersion estimates of each risk group should be considered under proper loss function. In this paper, by applying these two ideas, the benefit of the constrained Bayes estimates and balanced loss function will be discussed; in addition, application effectiveness will be proved through an analysis of real insurance accident data.

• Communications for Statistical Applications and Methods
• /
• 제27권1호
• /
• pp.97-108
• /
• 2020

This paper compares several methods for estimating parameters of normal inverse Gaussian distribution. Ordinary maximum likelihood estimation and the method of moment estimation often do not work properly due to restrictions on parameters. We examine the performance of adjusted estimation methods along with the ordinary maximum likelihood estimation and the method of moment estimation by simulation and real data application. We also see the effect of the initial value in estimation methods. The simulation results show that the ordinary maximum likelihood estimator is significantly affected by the initial value; in addition, the adjusted estimators have smaller root mean square error than ordinary estimators as well as less impact on the initial value. With real datasets, we obtain similar results to what we see in simulation studies. Based on the results of simulation and real data application, we suggest using adjusted maximum likelihood estimates with adjusted method of moment estimates as initial values to estimate the parameters of normal inverse Gaussian distribution.

• Communications for Statistical Applications and Methods
• /
• 제5권3호
• /
• pp.887-894
• /
• 1998

Let T=( $T_1$,…, $T_{k}$;k$\geq$2) be a minimal sufficient and complete statistic for a k-parameter exponential model. Consider a partition of T into ( $T_1$, $T_2$), where $T_1$=( $T_1$,…, $T_{r}$ and $T_2$=( $T_{r+1}$,…, $T_{k}$1$\leq$r$\leq$k-1/). This article represents a way to obtain higher moments such as skewness and kurtosis for the distribution T and the conditional distribution of $T_1$, given $T_2$= $t_2$. These results are illustrated by some examples.s.les.s.

• Communications for Statistical Applications and Methods
• /
• 제17권1호
• /
• pp.27-37
• /
• 2010

This article concerns the forecasting in binomial AR(p) models which is proposed by Wei$\ss$ (2009b) for time series of binomial counts. Our method extends to binomial AR(p) models a recent result by Jung and Tremayne (2006) for integer-valued autoregressive model of second order, INAR(2), with simple Poisson innovations. Forecasts are produced by conditional median which gives 'coherent' forecasts, and we estimate the forecast distributions of future values of binomial AR(p) models by means of a Monte Carlo method allowing for parameter uncertainty. Model parameters are estimated by the method of moments and estimated standard errors are calculated by means of block of block bootstrap. The method is fitted to log data set used in Wei$\ss$ (2009b).

• International Journal of Reliability and Applications
• /
• 제16권2호
• /
• pp.81-98
• /
• 2015

The Exponentiated Gamma (EG) distribution is one of the important families of distributions in lifetime tests. In this paper, a new generalized version of this distribution which is called kumaraswamy Exponentiated Gamma (KEG) distribution is introduced. A new distribution is more flexible and has some interesting properties. A comprehensive mathematical treatment of the KEG distribution is provided. We derive the $r^{th}$ moment and moment generating function of this distribution. Moreover, we discuss the maximum likelihood estimation of the distribution parameters. Finally, an application to real data sets is illustrated.

• Communications for Statistical Applications and Methods
• /
• 제16권2호
• /
• pp.389-396
• /
• 2009

Approximations for the null distribution of a test statistic arising in multivariate analysis to test homogeneity of variances and a mixture of two beta distributions by making use of a product of beta baseline density function and a polynomial adjustment, so called beta-polynomial density approximant, are discussed. Explicit representations of density and distribution approximants of interest in each case can easily be obtained. Beta-polynomial density approximants produce good approximation over the entire range of the test statistic and also accommodate even the bimodal distribution using an artificial example of a mixture of two beta distributions.

• Communications for Statistical Applications and Methods
• /
• 제20권4호
• /
• pp.321-327
• /
• 2013

Bayesian and empirical Bayesian methods have become quite popular in the theory and practice of statistics. However, the objective is to often produce an ensemble of parameter estimates as well as to produce the histogram of the estimates. For example, in insurance pricing, the accurate point estimates of risk for each group is necessary and also proper dispersion estimation should be considered. Well-known Bayes estimates (which is the posterior means under quadratic loss) are underdispersed as an estimate of the histogram of parameters. The adjustment of Bayes estimates to correct this problem is known as constrained Bayes estimators, which are matching the first two empirical moments. In this paper, we propose a way to apply the constrained Bayes estimators in insurance pricing, which is required to estimate accurately both location and dispersion. Also, the benefit of the constrained Bayes estimates will be discussed by analyzing real insurance accident data.

• Communications for Statistical Applications and Methods
• /
• 제24권4호
• /
• pp.397-419
• /
• 2017

A four-parameter extended fatigue lifetime model called the odd Birnbaum-Saunders geometric distribution is proposed. This model extends the odd Birnbaum-Saunders and Birnbaum-Saunders distributions. We derive some properties of the new distribution that include expressions for the ordinary moments and generating and quantile functions. The method of maximum likelihood and a Bayesian approach are adopted to estimate the model parameters; in addition, various simulations are performed for different parameter settings and sample sizes. We propose two new models with a cure rate called the odd Birnbaum-Saunders mixture and odd Birnbaum-Saunders geometric models by assuming that the number of competing causes for the event of interest has a geometric distribution. The applicability of the new models are illustrated by means of ethylene data and melanoma data with cure fraction.

• Journal of the Korean Statistical Society
• /
• 제35권3호
• /
• pp.281-303
• /
• 2006

In an earlier work, we investigated the problem of using linear programming to bound performance measures in a discrete time Jackson network. There it was assumed that the system evolution is controlled by the early arrival scheme. This assumption implies that the system can't be modelled by a Markov chain. This problem was resolved and performance bounds were calculated. In the present work, we use a modification of the early arrival scheme (without corrupting it) in order to make the system evolves as a Markov chain. This modification enables us to obtain explicit expressions for certain moments that could not be calculated explicitly in the pure early arrival scheme setting. Moreover, this feature implies a reduction in the linear program size as well as the computation time. In addition, we obtained tighter bounds than those appeared before due to the new setting.