Volume 9, Issue 1, Winter and Spring 2009, Page 1-256

The Prediction of the Maternal and Fetal Blood Lead Level via Generalized Linear Model

Zakaria Y. AL-Jammal

IRAQI JOURNAL OF STATISTICAL SCIENCES, 2009, Volume 9, Issue 1, Pages 1-16
DOI: 10.33899/iqjoss.2009.30637

Generalized linear models (GLMs) are generalization of the linear regression models, which allow fitting regression models to response variable that is non normal and follows a general exponential family. The aim of this study is to encourage and initiate the application of GLMs to predict the maternal and fetal blood lead level. The inverse Gaussian distribution with inverse quadratic link function is considered. Four main effects were significant in the prediction of the maternal blood lead level (pica, smoking of mother, dairy products intake of mother, calcium intake of mother), while in the prediction of the fetal blood lead level two main effects showed significance (dairy products intake of mother and hemoglobin of mother).
Keywords: Generalized linear models, Exponential family, Inverse Gaussian distribution, Link functions

Categorical data analysis with Practical Application

Furat B. Al-Dassy

IRAQI JOURNAL OF STATISTICAL SCIENCES, 2009, Volume 9, Issue 1, Pages 17-28
DOI: 10.33899/iqjoss.2009.30640

The main purpose of this study is to find out vectors that can be inserted to the statistical analysis of certain groups of variables which are formed as a result of a certain condition depending on categorical data (i.e. Qualitative variable ).
Many statistical models have been discussed to deal with this condition in order to have a discriminate function or to have variables by which we can specify the condition gradually. Depending on the importance of the inserted variable in that condition, a theorem form of this sensitive variable has been derived . Some applications have been also used in this paper.

Recognition of Musical ladders by Using Hidden Markov Model

IRAQI JOURNAL OF STATISTICAL SCIENCES, 2009, Volume 9, Issue 1, Pages 17-42
DOI: 10.33899/iqjoss.2009.30594

This research includes an idea that takes wide cares in modern applications through the object of Data Mining which is called Temporal Sequences. An investment of mathematical model known as Hidden Markov Model, of modeling Temporal Sequences. This research deals with an application combining between mathematics, computer and music. The problem of recognition of musical ladders is studied by using Hidden Markov Model, Some national songs are studied and modeled through this model. After making use of the computer for the songs under study, a result was reached to know a musical ladder which is used in national songs.

Using maximum distances from unit line to the Embedding vectors to estimate the delay time with an application

IRAQI JOURNAL OF STATISTICAL SCIENCES, 2009, Volume 9, Issue 1, Pages 43-62
DOI: 10.33899/iqjoss.2009.30596

In this research a new method was proposed to estimate the delay time. Which can be used to determine the relationship between the input and output of the series, this means that after each time interval, the input is affected by the output.
The advantages of this method are the simplicity and the basic calculations in linear and nonlinear transformations. The proposed technique was found by examining the greatest expansion from the unit line in the impacted space.

Using the Genetic Algorithm to solve some of the Inventory Models

IRAQI JOURNAL OF STATISTICAL SCIENCES, 2009, Volume 9, Issue 1, Pages 63-76
DOI: 10.33899/iqjoss.2009.30600

This research concentrates on the application of the Genetic Algorithm (GA) which is considered as an artificial search method on some of the Inventory Models. More than one which were suggested algorithm led to a number of solutions equal to the number of algorithm generation ,one of them was the optimal solution during a specified period of time .
Keywords :Genetic Algorithm;Optimization ;Inventory Model

A Comparison between the Prediction of State Space models and Stochastic Dynamic Linear Systems with Application

IRAQI JOURNAL OF STATISTICAL SCIENCES, 2009, Volume 9, Issue 1, Pages 77-92
DOI: 10.33899/iqjoss.2009.30612

This Paper is concerned with the synchronizing between numbers of dynamic linear system models with different parameters with their two different kinds. The first kind represented the equation error model which included ARMAX and ARX, also the second kind was the output error model which contained BJ and OE , and, another state space models, After the application of the above models on the real data ,the comparison has been done, we choose the ARX(1,7,3) and the state space model with 4 parameters which gave the minimum statistical criteria. These two models are also used in forecasting and another comparison has been done between two forecasted models, and we conclude that the forecasted dynamic linear system model is the best which gave the minimum statistical criteria.

Stochastic zero – one programming

IRAQI JOURNAL OF STATISTICAL SCIENCES, 2009, Volume 9, Issue 1, Pages 107-126
DOI: 10.33899/iqjoss.2009.30618

zero - one programming case from integer linear programming where the variable's are equal to zero or one, the decision factor uses this kind from programming when he meets him problems of the kind yes or no.
The stochastic zero-one programming construction formed is used when one or all parameters of model(cj,bi,aij) are random variable taken mathematical distribution.
In this research we discuss stochastic zero-one programming problem where (aij) random variable (construction and solution) and use it in practical application on some vegetative crops in Iraq.

Fuzzy Autoregressive Model With an Application

IRAQI JOURNAL OF STATISTICAL SCIENCES, 2009, Volume 9, Issue 1, Pages 127-146
DOI: 10.33899/iqjoss.2009.30623

The research is dedicated in the study to time series and the ability of using fuzzy logic with it in order to develop forecasting approaches. In this research ,the time series (Autoregressive model) are linked with fuzzy logic in order to get on the Parameters of fuzzy time series models (fuzzy logic Autoregressive model),and applied that on the data of daily mistakes rates in charges production. The fuzzy Autoregressive model of time series gave forecasting more sutiable than those given by Autoregressive model.

A new Separated Ratio Estimator in the stratified random sampling

IRAQI JOURNAL OF STATISTICAL SCIENCES, 2009, Volume 9, Issue 1, Pages 147-160
DOI: 10.33899/iqjoss.2009.30627

The aim of this research is to find a new separated ratio estimator in the stratified random sampling ,three other estimators are compared with the new separated ratio estimator , namely : ordinary separated ratio estimator , combined ratio estimator, and the suggested ratio estimator by ( Kadilar & Cingi ; 2005 ) . Using the (MSE) as a comparative criterion , we proved that our proposal estimator has more precision than the others . Many numerical and simulation examples have been applied .
Keywords :Stratified Random Sampling ;Ratio–Estimator; Mean Square Errors

Types of Unusual Observations in Multiple Regression & Some Methods of it’s Diagnostic with Application

IRAQI JOURNAL OF STATISTICAL SCIENCES, 2009, Volume 9, Issue 1, Pages 161-192
DOI: 10.33899/iqjoss.2009.30631

The work in this paper is a diagnosis of three types of unusual observations in multiple regression, the outlier observation is diagnostic by using Boxplot & studentized residuals, Diagnostic leverage points by using (Hat matrix) , & diagnostic of the influence observations by using (dfbeta).
In practice the work is comparing the effects of omitting outlier observations in the normal distribution of the residuals to the equation which is building to the Thalassaemia disease and Treating the multicollinearity by omitting some variables and using ridge regression, getting a good model agrees with the viewing of medicine by using (SAS.9) package.

Using the Method of Global Criterion in Multi-Objective Mathematical Programming

IRAQI JOURNAL OF STATISTICAL SCIENCES, 2009, Volume 9, Issue 1, Pages 193-226
DOI: 10.33899/iqjoss.2009.30633

The research deals with the most important specific development of the traditional mathematical programming (MP) and shows the mathematical programming with multi-objective (MOMP) which forms the vertebral column in application of Multi-Criteria Decision Making (MCDM), Decision Support System (DSS), operations research / management science (OR/MS).
After studying these new models and tools and their uses, the researcher chose Global Criterion Method to study it concept and properties, limitation and stages of the solution to achieve the algorithm and the flow-chart of it, and using the Global Criterion Method to obtain the best feasible final solutions to constrained decision, linear, multiple objective problem without any priority or weighted.
Finally, the study concludes too many feasible solutions which are called, non-dominated solutions to the application of a case study and comparison with the results of using other models.

"The Two Methods of Additive Holt Winters and Fuzzy Logic in Predicting the Time Series"

IRAQI JOURNAL OF STATISTICAL SCIENCES, 2009, Volume 9, Issue 1, Pages 237-256
DOI: 10.33899/iqjoss.2009.30636

The prediction of future behavior of this under study phenomenon is considered as the most important and vital subject in statistics. This study took a great deal of concern by statistics' researchers in order to estimate the parameters of time series' samples, consequently, using it in the prediction and controlling of areas for applicatory fields.
Recently, many researches appeared to use modern computer technologies in analyzing nonlinear series, besides, many different statistical subjects, such as neural-networks and fuzzy logic.
The purpose of this research is characterized by using fuzzy logic method and comparing it with Holt Winters additive method depending upon mean square error MSE to predict the chronological series' values. The fuzzy logic method had overcome the additive Holt Winters method according to the statistical norms. Therefore, the fuzzy method could be the best and the most accurate method in predicting the time series.