|
|
|
 |
Search published articles |
 |
|
Showing 27 results for Bayes
, ,
Volume 17, Issue 1 (9-2012)
Abstract
In many experiments about lifetime examination, we will faced on restrictions of time and sample size, which this factors cause that the researcher can’t access to all of data. Therefore, it is valuable to study prediction of unobserved values based on information of available data. in this paper we have studied the prediction of unobserved values in two status of one-sample and two-sample, when the parent distribution is the exponential distribution and imposed restriction is double censoring. in each case the interval prediction by given cover will be obtain. Finally, a numerical example is given to illustrate the procedures.
Mr Saeed Bagrezaei, Mr Ebrahim Aminiseresht,
Volume 18, Issue 2 (3-2014)
Abstract
According to the first nth observations of the upper record from exponential distribution, in this article, we can compute maximum likelihood estimation of this distribution parameter. We, then, concentrate on point prediction of the future upper record values in exponential distribution based both on classic and Bayes approaches and second degree and linex loss functions.We, ultimately, deal with numerical comparison available point predictions through Monte Carlo simulation.
Atefe Mokhtari Hasanabadi, Manouchehr Kheradmandnia,
Volume 18, Issue 2 (3-2014)
Abstract
Recently several
control charts have been introduced in the statistical process control literature which are based on the idea of Bayesian Predictive Density
(BPD). Among these charts is the
variation control chart which we refer to it as VBPD chart. In this paper we add the idea of Moving Average to VBPD
chart and introduce a new variation control chart which has all advantages of the original VBPD
chart and in addition has a new advantage which is its sensitivity to small changes in process variance. We
refer to this new chart as MAVBPD chart. In both VBPD and MAVBP charts , the parameters are assumed unknown
but the control statistic has a known F distribution which means that, the
control limits can be obtained without
simulation.
Student Atefe Javidi, Student Somayeh Rahpeima, Dr Majid Jafari Khaledi,
Volume 18, Issue 2 (3-2014)
Abstract
Statistical models are utilized to learn about the mechanism that the data are generating from it. Often it is assumed that the random variables y_i,i=1,…,n ,are samples from the probability distribution F which is belong to a parametric distributions class. However, in practice, a parametric model may be inappropriate to describe the data. In this settings, the parametric assumption could be relaxed and more flexible models could be used analysis of data. In the nonparametric Bayes approach, a prior distributions is defined over the whole space of probability distributions for random variable distribution. Due to the Dirichlet process (DP) has interesting properties, it is thus used extensively. In this paper, we introduce DP and its features.
Mahsa Abedini, Iraj Kazemi,
Volume 19, Issue 1 (6-2014)
Abstract
In previous studies on fitting non-linear regression models with the symmetric structure the normality is usually assumed in the analysis of data. This choice may be inappropriate when the distribution of residual terms is asymmetric. Recently, the family of scale-mixture of skew-normal distributions is the main concern of many researchers. This family includes several skewed and heavy-tailed distributions, such as the skew-t and the skew slash, as special cases and is recommended as an alternative to the normal distribution. The statistical inference based on the maximization of marginal likelihoods is complicated, in general, for non-linear regression models and thus we implement the MCMC approach to obtain Bayes estimates. Finally, we fit a non-linear regression model using proposed distributions for a real data set to show the importance of the recommended model.
Mehrangiz Falahati-Naeini,
Volume 19, Issue 1 (6-2014)
Abstract
In this article introduce the sequential order statistics. Therefore based on multiply Type-II censored sample of sequential order statistics, Bayesian estimators are derived for the parameters of one- and two- parameter exponential distributions under the assumption that the prior distribution is given by an inverse gamma distribution and the Bayes estimator with respect to squared error loss is calculated. Moreover, prediction of future failure time is considered.
Finally in example Bayesian estimator and non-bayesian estimatores, namely the Best Linear Unbiased Estimator (BLUE) and Approximate Maximum Likelihood Estimator (AMLE) are derived.
Fahimeh Moradi, Ali Karimnezhad, Soodabeh Shemehsavar,
Volume 19, Issue 1 (6-2014)
Abstract
Bayesian networks (BNs) are modern tools for modeling phenomena in dynamic and static systems and are used in different subjects such as disease diagnosis, weather forecasting, decision making and clustering. A BN is a graphical-probabilistic model which represents causal relations among random variables and consists of a directed acyclic graph and a set of conditional probabilities. Structure learning and parameter learning are two main subjects in BNs. In this paper, we consider a BN with a known structure and then, by simulate some data, we try to learn structure of the network using two well-known algorithms, namely, PC and $ K_{2} $ algorithms. Then, we learn parameters of the network and derive the maximum likelihood, maximum a posteriori and posterior mean estimates of the corresponding parameters. Furthermore, we compare performance of the estimates using the Kullback-Leibler divergence criteria and finally, utilizing a real data set, we consider the structure and parameter learning tasks to illustrate practical utility of the proposed methods.
, , ,
Volume 20, Issue 1 (4-2015)
Abstract
The problem of sample size estimation is important in medical applications, especially in cases of expensive measurements
of immune biomarkers. This paper describes the problem of logistic regression analysis with the sample
size determination algorithms, namely the methods of univariate statistics, logistics regression, cross-validation and
Bayesian inference. The authors, treating the regression model parameters as multivariate variable, propose to estimate
the sample size using the distance between parameter distribution functions on cross-validated data sets.
Herewith, the authors give a new contribution to data mining and statistical learning, supported by applied mathematics.
Dr farzad Eskandari, Ms imaneh Khodayari Samghabadi,
Volume 21, Issue 1 (9-2016)
Abstract
There are different types of classification methods for classifying the certain data. All the time the value of the variables is not certain and they may belong to the interval that is called uncertain data. In recent years, by assuming the distribution of the uncertain data is normal, there are several estimation for the mean and variance of this distribution. In this paper, we consider the mean and variance for each of the start and end of intervals. Thus we assume that the distribution of uncertain data is bivariate normal distribution. We used the maximum likelihood to estimate the means and variances of the bivariate normal distribution. Finally, Based on the Naive Bayesian classification, we propose a Bayesian mixture algorithm for classifying the certain and uncertain data. The experimental results show that the proposed algorithm has high accuracy.
Dr Vahid Rezaeitabar, Selva Salimi,
Volume 21, Issue 1 (9-2016)
Abstract
A Bayesian network is a graphical model that represents a set of random variables and their causal relationship via a Directed Acyclic Graph (DAG). There are basically two methods used for learning Bayesian network: parameter-learning and structure-learning. One of the most effective structure-learning methods is K2 algorithm. Because the performance of the K2 algorithm depends on node ordering, more effective node ordering inference methods are needed. In this paper, based on the fact that the parent and child variables are identified by estimated Markov Blanket (MB), we first estimate the MB of a variable using Grow-Shrink algorithm, then determine the candidate parents of a variable by evaluating the conditional frequencies using Dirichlet probability density function. Then the candidate parents are used as input for the K2 algorithm. Experimental results for most of the datasets indicate that our proposed method significantly outperforms previous method.
Ali Aghmohammadi, Sakine Mohammadi,
Volume 21, Issue 2 (3-2017)
Abstract
Dynamic panel data models include the important part of medicine, social and economic studies. Existence of the lagged dependent variable as an explanatory variable is a sensible trait of these models. The estimation problem of these models arises from the correlation between the lagged depended variable and the current disturbance. Recently, quantile regression to analyze dynamic panel data has been taken in to consideration. In this paper, quantile regression model by adding an adaptive Lasso penalty term to the random effects for dynamic panel data is introduced by assuming correlation between the random effects and initial observations. Also, this model is illustrated by assuming that the random effects and initial values are independent. These two models are analyzed from a Bayesian point of view. Since, in these models posterior distributions of the parameters are not in explicit form, the full conditional posterior distributions of the parameters are calculated and the Gibbs sampling algorithm is used to deduction. To compare the performance of the proposed method with the conventional methods, a simulation study was conducted and at the end, applications to a real data set are illustrated.
Shahrastani Shahram Yaghoobzadeh,
Volume 21, Issue 2 (3-2017)
Abstract
In this study, E-Bayesian of parameters of two parameter exponential distribution under squared error loss function is obtained. The estimated and the efficiency of the proposed method has been compared with Bayesian estimator using Monte Carlo simulation.
, ,
Volume 22, Issue 1 (12-2017)
Abstract
Analysis of large geostatistical data sets, usually, entail the expensive matrix computations. This problem creates challenges in implementing statistical inferences of traditional Bayesian models. In addition,researchers often face with multiple spatial data sets with complex spatial dependence structures that their analysis is difficult. This is a problem for MCMC sampling algorithms that are commonly used in Bayesian analysis of spatial models, causing serious problems such as slowing down and chain integration. To escape from such computational problems, we use low-rank models, to analyze Gaussian geostatistical data. This models improve MCMC sampler convergence rate and decrease sampler run-time by reducing parameter space. The idea here is to assume, quite reasonably, that the spatial information available from the entire set of observed locations can be summarized in terms of a smaller, but representative, sets of locations, or ‘knots’. That is, we still use all of the data but we represent the spatial structure through a dimension reduction. So, again, in implementing the reduction, we need to design the knots. Consideration of this issue forms the balance of the article. To evaluate the performance of this class of models, we conduct a simulation study as well as analysis of a real data set regarding the quality of underground mineral water of a large area in Golestan province, Iran.
Dr Fatemeh Hosseini, Dr Omid Karimi, Ms Ahdiyeh Azizi,
Volume 23, Issue 1 (9-2018)
Abstract
Often in practice the data on the mortality of a living unit correlation is due to the location of the observations in the study. One of the most important issues in the analysis of survival data with spatial dependence, is estimation of the parameters and prediction of the unknown values in known sites based on observations vector. In this paper to analyze this type of survival, Cox regression model with piecewise exponential function used as a hazard and spatial dependence as a Gaussian random field and as a latent variable is added to the model. Because there is no closed form for posterior distribution and full conditional distributions, also long computing for Markov chain Monte Carlo algorithms, to analyze the model are used the approximate Bayesian methods.
A practical example of how to implement an approximate Bayesian approach is presented.
Aliakbar Rasekhi,
Volume 23, Issue 2 (3-2019)
Abstract
WinBUGS is one of the usual softwares in computational Bayesian statistics, which is used to fit Baysian models easily. Although this software has usual mathematical functions and statistical distributions as built in functions, sometimes it is necessary to include other functions and distributions in computations which is done by some tricks and indirectly. By using WinBUGS development interface (known as WBDev), new mathematical functions and statistical distributions can be added in the software. This method facilitates writing codes of statistical models, increases speed of computations and make computations more efficient. In this paper, the stages of including new mathematical functions and statistical distributions in the WinBUGS are illustrated by some examples.
Ali Shadrokh, Shahrastani Shahram Yaghoobzadeh,
Volume 24, Issue 1 (9-2019)
Abstract
In this study, E-Bayesian and hierarchical Bayesian of parameter of Rayleigh distribution under progressive type-II censoring sampales and the efficiency of the proposed methods has been compared with each and Bayesian estimator using Monte Carlo simulation.
Shahrastani Shahram Yaghoobzadeh,
Volume 24, Issue 1 (9-2019)
Abstract
In this paper, reliability in multi-component stress-strength models, when the stress and strength variables are inverse Rayleigh distributions with different parameters of alpha and beta. Estimates of the maximum likelihood, Bayesian and empirical Bayesian are estimated. Then, with the help of Monte Carlo simulation and two real data sets, these estimation methods are compared.
Dr Fatemeh Hosseini, Dr Omid Karimi, Miss Fatemeh Hamedi,
Volume 24, Issue 1 (9-2019)
Abstract
Tree models represent a new and innovative way of analyzing large data sets by dividing predictor space into simpler areas. Bayesian Additive Regression Trees model, a model that we explain in this article, uses a totality of trees in its structure, since the combination of several trees from a tree only has a higher accuracy.
Then, this model is a tree-based model and a nonparametric model that uses general aggregation methods, and boosting algorithms in particular and in fact is extension of the classification and Regression Tree methods in which the decision tree exists in the structure of these methods.
In this method, on the parameters of the model sum of tree and put regular prior then use the boosting algorithms for analysis. In this paper, first the Bayesian Additive Regression Trees model is introduced and then applied in survival analysis of lung cancer patients.
, , ,
Volume 24, Issue 2 (3-2020)
Abstract
In the analysis of Bernoulli's variables, an investigation of the their dependence is of the prime importance. In this paper, the distribution of the Markov logarithmic series is introduced by the execution of the first-order dependence among Bernoulli variables. In order to estimate the parameters of this distribution, maximum likelihood, moment, Bayesian and also a new method which called the expected Bayesian method (E-Bayesian) are employed. In continuation, using a simulation study, it is shown that the expected Bayesian estimator out performed over the other estimators.
Dr. Shahram Yaghoobzadeh Shahrestani, Dr. Reza Zarei,
Volume 25, Issue 1 (1-2021)
Abstract
Whenever approximate and initial information about the unknown parameter of a distribution is available, the shrinkage estimation method can be used to estimate it. In this paper, first, the E-Bayesian estimation of the parameter of an inverse Rayleigh distribution under the general entropy loss function is obtained. Then, the shrinkage estimate of the inverse Rayleigh distribution parameter is investigated using the guess value. Also, using Monte Carlo simulations and a real data set, the proposed shrinkage estimation is compared with the UMVU and E-Bayesian estimators based on the relative efficiency criterion.
|
|
