Showing papers in "Journal of Statistics Applications & Probability in 2015"

The Generalized Inverted Generalized Exponential Distribution with an Application to a Censored Data

Abstract: We propose a two parameter Inverted Generalized Exponential (IGE) and a three parameter Generalized Inverted Generalized Exponential (GIGE) probability models as generalizations of the one-parameter Exponential distribution and some other distributions in the literature. We explore the statistical properties of the GIGE distribution and its parameters were estimated for both censored and uncensored cases using the method of maximum likelihood estimation (MLE). An application to a real data set is also provided to assess the flexibility of the GIGE distribution over some of its sub-models.

Bayesian Estimation and Prediction for Pareto Distribution based on Ranked set Sampling

Abstract: In this paper, we provide Bayesian estimation for the parameters of the Pareto distribution based on simple random sample (SRS) and ranked set sampling (RSS) in two cases, one cycle and m-cycle. Posterior risk of the derived estimators are also ob tained by using squared error loss (SEL). Two-sample Bayesian prediction for future observations are obtained by using SRS and RSS in two cases, one cycle and m-cycle. A simulation data for SRS and RSS for one cycle and two cycle are used to illustrate the results.

On The Estimation of Population Mean in Current Occasion in Two- Occasion Rotation Patterns

Abstract: The present work is an effort to develop some estimators of current population mean in two-occasion successive sampling utilizing the known population mean

A Merton Model of Credit Risk with Jumps

Abstract: In this note we consider a Merton model for default risk, wher e the firm’s value is driven by a Brownian motion and a compound Poisson process.

Linear Models for S-Shaped Growth Curves

Abstract: The S-shaped or sigmoidal models have two important members as Logistic model and Gompertz model. These models are very useful in studies of medical science, actuarial scienc e and biological sciences. The proposed linear models are the alternative of Logistic model and Gompertz model respectively. Linear models satisfy the property of best linear unbiased estimator ( BLUE) and removed the problem of nonlinear least squares estimation mentioned by Ratkowsky (1983, 1989) and Bates & Watts (1980). The goodness of fit for the proposed linear models have been verifi ed with the help of several published data sets.

On Stochastic Multi-Level Multi-Objective Fractional Programming Problems

Abstract: This paper proposes an approach where it can be applied to the optimization decisions making problems under uncertainties and solves a multi-level multi-objective fr actional programming problems involving stochastic parameters coefficient in objective functions (SMLMOFPP). In this work, the first phas e of the solution approach, we convert the probabilistic nat ure (stochastic) of this problem in objective functions into a m ulti-level multi-objective fractional programming probl ems (MLMOFPP).At the second phase, we use a computer-oriented technique to convert (MLMOFPP) into a multi-level multi-objective linear programming problems (MLMOLPP). Then a fuzzy approach solves (MLMOLPP) using the concept of tolerance membership function to develop a Tchebycheff problem for generating a compromise solution for this problem. In addition, a numerical example is provided to demonstrate the correctness of the proposed solution.

Bayesian Estimation of Exponentiated Gamma Parameter for Progressive Type II Censored Data With Binomial Removals

Abstract: In the present study, we have considered the exponentiated gamma distribution as an important life time model for the situations where hazard rate function is either monotonic i ncreasing or in the bathtub shape. We propose Bayes estimators of the parameter of the exponentiated gamma distribution under general entropy loss function, squared error loss function an d we have also derived its maximum likelihood estimator. The estimators have been compared through their simulated risks.

A Comparative Study based on Bayes Estimation under Progressively Censored Rayleigh Data

Abstract: A comparative study based on two different asymmetric loss functions presented in this article. Two-parameter Rayleigh model is consider here as the underline model for the present comparative study, that evaluate the properties of Bayes estimators under progressive Type-II right censored data.

On some Classical Properties of the Mixture of Burr XII and Lomax Distributions

Abstract: The BurrXII and Lomax distributions are the most widely and important distributions used for life time purpose and for modeling the business failure time data. Burr XII distribution is mainly used to explain the allocation of wealth and wages among the people of the particular society. And Lomax is used to model the business failure time data. It is a transformed shape of Pareto distribution. In this research paper, the properties of the mixture of burr XII and Lomax distributions have been given. Classical properties such as Cumulative distribution function, failure rate, hazard rate, odd function, inverse hazard function and the cumulative hazard function, , r th moment, , moments, mean and variance have been derived.

An improved Exponential-Type Estimator for Finite Population Variance Using Transformation of Variables

Abstract: In this article we have proposed an improved exponential-type for estimating the unknown population variance of the study variable, using transformation on both the study as well as on the auxiliary variable. The properties of the sugges ted estimator have been discussed up to first order of approximation. We als o have derived some efficiency comparison conditions under w hich the proposed variance estimator performed better than the usual unbiased estimator, traditional ratio estimator and ( 1) estimators. Theoretical efficiency conditions are verified numerically by taking some real data sets taken from the literature, with a result in increasing the efficiency.

Inference of Progressively Censored Data from The Generalized Exponential Distribution

Abstract: In this paper, we derive approximate moments of progressively type-II right censored order statistics from the general ized exponential distribution . Also, using these moments to derive the best linear unbiased estimates and maximum likelihood estimates of the location and scale parameters from the generalized exponential distribution. In addition, we use Monte-Carlo simulation method to obtain the mean square error of the best linear unbiased estimates and maximum likelihood estimates and make comparison between them. Finally, we will present numerical example to illustrate the inference procedures developed in this distributio n.

Simulation-based Analysis of Degradation Data: Implementation of a Multilevel Approach

Abstract: Data often arrive with hierarchical structure and multilevel regression modeling is the most popular approach to handle such data. This paper demonstrates how multilevel model can be analyzed in Bayesian framework, with reference to a practical degradation data problem. Assuming a varying-intercept, varying-slope model for the data, the exact as well as the approximate inference procedures have been developed using R and JAGS and their performance have been compared. Further, the concept of Bayesian p-values have been discussed to assess the adequacy of the proposed model.