|
|
|
 |
Search published articles |
 |
|
Showing 295 results for Type of Study: Applicable
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.
Shima Hajizadeh, Majid Sarmad,
Volume 19, Issue 2 (2-2015)
Abstract
In many diverse scientific fields, the measurements are directions. For instance, a biologist may be measuring the
direction of flight of a bird or the orientation of an animal. A series of such observations is called ”directional
data”. Since a direction has no magnitude, these can be conveniently represented as points on the circumference of
a unit circle centered at the origin or as unit vectors connecting the origin to these points. Because of this circular
representation, such observations are also called circular data. In this paper, circular data will be introduced at first
and then it is explained how to calculate the mean direction, dispersion and higher moments. The solutions to many
directional data problems are often not obtainable in simple closed analytical forms. Therefore, computer softwares
is essential to use these methods. At the end of this paper, the CircStat’s package has been used to analyze data sets
in R and Matlab softwares.
Faegheh Amiri, manouchehr kheradmandnia,
Volume 19, Issue 2 (2-2015)
Abstract
In many quality control applications, the necessary distributional assumptions to correctly apply the traditional parametric control charts are either not met or there is simply not enough information or evidence to verify the assumptions. It is well known that performance of many parametric control charts can be seriously degraded in situations like this. Thus, control charts that do not require a specific distributional assumption to be valid, so-called nonparametric or distribution-free charts, are desirable in practice. In this paper, a simple to use multivariate nonparametric control chart is introduced. The chart is based on the multivariate two sample Mann-Withney Wilcoxon test for equality of location vectors of two populations. Using simulated data we show that there are situations in which the Mann-Withney multivariate control chart has a better performance compared with T2 control chart.
Sakineh Dehghan, Mohammadreza Farid-Rohani,
Volume 20, Issue 1 (4-2015)
Abstract
In this article, first, we introduce depth function as a function for center-outward ranking. Then we present and
use half space or Tukey depth function as one of the most popular depth functions. In the following, multivariate
nonparametric tests for location and scale difference between two population are expressed by ranking and statistics
based on depth versus depth plot. Finally, according to these tests, performance of the suggested non-invasive
distraction method for pain intensity, life quality, operative ability and inflation rate is evaluated for osteoarthritic
and is compared with usual invasive distraction method.
Mrs Zahra Niknam, Dr mohammad Hossein Alamatsaz,
Volume 20, Issue 1 (4-2015)
Abstract
In many issues of statistical modeling, the common assumption is that observations are normally distributed. In
many real data applications, however, the true distribution is deviated from the normal. Thus, the main concern of
most recent studies on analyzing data is to construct and the use of alternative distributions. In this regard, new
classes of distributions such as slash and skew-slash family of distributions have been introduced .This has been the
main concern of many researcher’s investigations in recent decades. Slash distribution, as a heavy tailed symmetric
distribution, is known in robust studies. But since , in empirical examples, there are many situations where symmetric
distributions are not suitable for fitting the data study of skew distributions has become of particular importance.In
this paper we introduce skew-slash distribution and study their properties. Finally, some applications to several real
data sets are illustrated in order to show the importance of the distribution in regression models.
,
Volume 20, Issue 2 (10-2015)
Abstract
Methods for small area estimation have been received great attention in recent years due to growing demand for
reliable small area estimation that are needed in development planings, allocation of government funds and marking
business decisions. The key question in small area estimation is how to obtain reliable estimations when sample
size is small. When only a few observations(or even no observation) are available from a given small area, small
sample sizes lead to undesirably large standard errors. The only possible solution to the estimation problem is to
borrow strength from available data sets. This is accomplish by using appropriate linking models (included explicit
and implicit models) to increas the effect of sample size for estimation. The generalized linear mixed models and
the empirical best linear unbiased predictor, are extensively used to estimate reliable mean of small areas. In this
article,first we introduce the small area estimation.Then, to obtain reliable small area estimations we introduce the
Fay-Herriot model as a special case of the generalized linear mixed model. Finally, in an Simulation study we use
Iran 1382 agricultural census data to estimate orange production in Fars cities (small areas) in the year 1382 based
on Fay-Herriot model.
Shirin Shahsanam, Masoud Yarmohammadi,
Volume 20, Issue 2 (10-2015)
Abstract
Nowadays factor analysis has been extended greatly, and one of its applications is to analysis the attributes which are
not measurable directly. While the response variable has a Bernoulli distribution, using factor analysis method for
continuous quantities leads to invalid and misleading results. Factor analysis for Bernoulli response variable base on
Logit model is developed in this paper and its applications have been explained for read data from a research project
about the high school mathematics textbooks in Iran.
Anita Abdollahi,
Volume 20, Issue 2 (10-2015)
Abstract
In this paper, after stating the characteristicof some of continuous distributions including, gamma, Crovelli’s
gamma, Rayleigh, Weibull, Pareto, exponential and generalized gamma distribution with each other,these distributions
were fit on drought data of Guilan state and the best distribution was presented. Then, severity and duration of
the drought of different sites were investigated using standardized precipitation index.
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.
Mr Majid Janfada, Dr Davood Shahsavani,
Volume 21, Issue 2 (3-2017)
Abstract
The study of many scientific and natural phenomena in laboratory condition is sometimes impossible, therefore theire expresed by mathemathical models and simulated by complex computer models (codes). Running a computer model with different inputs is called a computer expriment.
Statistical issues allocated a wide range of applications for computer expriment to itself. In this paper, the structure of computer models is described, and one of statistical applications, that is variance-based sensitivity analysis is expressed. Sensitivity analysis, involves a set of methods that determine the effect on model inputs on the output by using sensitivity indices. The indices are defined based on the concept of condition variance and the since explicit mathematical form of the model is unclear, hence the essues monte carlo based them are proposed.
Due to the inherent complexity of the model, execuation time is problem.Therefore a specifict design of expriment, base on Quasi-random number, is proposed to reduce the runnig costs. As an application, the INCA-N model that simulates amount of Nitrogen in river and underground sources was used. Using the sensitivity indices, we could found the effective variable on this danger pollution that threaten human life and inviromental.
, ,
Volume 21, Issue 2 (3-2017)
Abstract
Copula functions as a model can show the relationship between variables. Appropriate copula function for a specific application is a function that shows the dependency between data in a best way. Goodness of fit tests theoretically are the best way in selection of copula function. Different ways of goodness of fit for copula exist. In this paper we will examine the goodness of fit tests from theoretical point of view and evaluate three different methods for comparing the copula functions as well as numerical comparison in order to show the advantage and weak points of each method. At the end we will analyze the methods of discussed test by using the information from Tehran Stock Exchange.
Maryam Shekarisaz, Hamidreza Navvabpour,
Volume 21, Issue 2 (3-2017)
Abstract
In many statistical studies some units do not respond to a number or all of the questions. This situation causes a problem called non-response. Bias and variance inflation are two important consequences of non-response in surveys. Although increasing the sample size can prevented variance inflation, but cannot necessary adjust for the non-response bias. Therefore a number of methods are used for reducing non-response effects. In the cases where missing mechanism is at random, weighting adjustment is an appropriate method for compensating the effects of unit non-response. Propensity score is a weighting method in which weight allocation is accomplished based on the estimates of response probabilities. These estimates are obtained by fitting suitable parametric models. In this paper, the propensity score method and its resulted adjusted estimators are introduced. Then we compare the performance of three propensity score adjusted estimators. Finally, data on Household Income and Expenditure Survey for urban families conducted by Statistical Centre of Iran in spring 1390 are used to compare the adjusted propensity score estimators by two measures of comparisons, root relative mean squared error and relative efficiency.
, ,
Volume 21, Issue 2 (3-2017)
Abstract
Analytic combinatorics aims to enable precise quantitative predictions of the properties of large combinatorial structures. This theory has emerged over recent decades as essential both for the analysis of algorithms and for the study of scientific models in many disciplines, including probability theory, statistical physics, computational biology and information theory. With a careful combination of symbolic enumeration methods, complex analysis, generating functions and saddle point analysis, it can be applied to study of fundamental structures such as permutations, sequences, strings, walks, paths, trees, graphs and maps. This paper aims to introduce the order steps of an analytic combinatorics.
Majid Abiar, Abdolrahim Badamchizadeh,
Volume 22, Issue 1 (12-2017)
Abstract
In this paper, an M/M/1 queue with instantaneous Bernoulli feedback is studied in the event of server failure, the catastrophe occurs and after repair, it starts to work again. The transient response for the probability function of the system size is presented. The steady state analysis of system size probabilities and some performance measures of system are provided. Then the results are used to consider the performance of an ATM. Then to observe and optimize the performace of the ATM, we illustrate the effects of changing parameters on system performance measures. At last, we simulate the system by using the R application. Then we compare its results with expected results.
, , ,
Volume 22, Issue 1 (12-2017)
Abstract
Latent class analysis (LCA) is a method of evaluating non sampling errors, especially measurement error in categorical data. Biemer (2011) introduced four latent class modeling approaches: probability model parameterization, log linear model, modified path model, and graphical model using path diagrams. These models are interchangeable. Latent class probability models express likelihood of cross-classification tables in term of conditional and marginal probabilities for each cell. In this approach model parameters are estimated using EM algorithm. To test latent class model chi-square statistic is used as a measure of goodness-of-fit. In this paper we use LCA and data from a small-scale survey to estimate misclassification error (as a measurement error) of students who had at least a failing grade as well as misclassification error of students with average grades below 14.
, ,
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.
,
Volume 22, Issue 2 (3-2018)
Abstract
Anita Abdollahi Nanvapisheh,
Volume 22, Issue 2 (3-2018)
Abstract
In this paper, first, we investigate probability density function and the failure rate function of some families of exponential distributions. Then we present their features such as expectation, variance, moments and maximum likelihood estimation and we identify the most flexible distributions according to the figure of probability density function and the failure rate function and finally we offer practical examples of them.
Ms Sara Jazan, Dr Seyyed Morteza Amini,
Volume 22, Issue 2 (3-2018)
Abstract
One of the factors affecting the statistical analysis of the data is the presence of outliers. The methods which are not affected by the outliers are called robust methods. Robust regression methods are robust estimation methods of regression model parameters in the presence of outliers. Besides outliers, the linear dependency of regressor variables, which is called multicollinearity, the large number of regressor variables with respect to sample size, specially in high dimensional sparse models, are problems which result in efficiency reduction of inferences in classical regression methods. In this paper, we first study the disadvantages of classical least squares regression method, when facing with outliers, multicollinearity and sparse models. Then, we introduce and study robust and penalized regression methods, as a solution to overcome these problems. Furthermore, considering outliers and multicollinearity or sparse models, simultaneously, we study penalized-robust regression methods. We examine the performance of different estimators introdused in this paper, through three different simulation studies. A real data set is also analyzed using the proposed methods.
Dr Ehsan Bahrami Samani,
Volume 23, Issue 1 (9-2018)
Abstract
In this paper, we propose Hurdle regression models for analysing count responses with extra zeros. A method of estimating maximum likelihood is used to estimate model parameters. The application of the proposed model is presented in insurance dataset. In this example, there are many numbers of claims equal to zero is considered that clarify the application of the model with a zero-inflated count response. Different count regression models are introduced in this paper to model such data sets. Including Hurdle Poisson and Hurdle Negative Binomial regression models.
|
|
