IRAOI JOURNAL OF STATISTICAL SCIENCES

This research consists of using some statistical techniques to study
time series for universal prices of wheat as output series and universal
product of wheat as input series. By using transfer function on stationary
data first , and secondly on stationary fuzzy data , then compare between
these two cases to obtain the best transfer function model for data through
forecasting criteria to comparing between these two cases .The most
suitable model for this data was the transfer function model for stationary
fuzzy data because it has minimum value for forecasting criteria .

The regression control chart is an effective statistical process control
tool in situations where the output characteristic of interest is affected by
an external covariate . In this research , a regression control chart has been
formed from data obtained from the Badush cement plant (expansion). The
data are production volume of cement (dependent variable y) which is
affected by the manhours used in the plant (independent variable x) and the
data were taken during the time period (2009-2010) . For purposes of
statement performance of the plant, we applied the regression control chart
on data of a later time period (2011) and there was one point out of upper
control limit and this means that there was a significant improvement in the
level of performance and an increase in the rate of productivity,
(temporarily or permanently) .As a result an investigation was made to
know the reason which led to the exit point for the upper control limit and
approval for the development of the plant and increase of the production.

This research deals with the study of the stochastic weibull process
in two cases of incomplete data: the failure truncated case and the time
truncated case. These parameters were estimated using parametric method
by the maximum likelihood method, to find the rate of occurrence failure
times. The kernel estimator was used as a non-parametric method for
estimating the rate of occurrence and comparison between two estimators.
The unbiased estimator and confidence limits for the shape parameter were
found in both cases.

This study includes practical application on stops times (hours) in
oven the Badush Cement plant , as time rate is estimated to occur stops the
oven and find kernel estimator and a comparative study of the results in
both cases. It is clear by comparing the efficiency and accuracy of
estimating the parameters in the time truncated case best of failure
truncated case , and that the time rate for occur stops oven continues to
increase over time.

It is well known that employing the overlapping and interlocking
relations among multiple phenomena offers modeling that is more accurate
than these which are obtained by analyze each phenomena separately. In
this research interrelationship for three phenomena that appear in the
Meteorology analyzed which are monthly averages for the maximum
temperatures, monthly averages for minimum temperatures and monthly
averages for evaporation, through a system of three equations by using the
Vector ARMA Model. For the purpose of analysis, we used monthly data
for the period from January 1981 to December 2010 while keeping the last
ten months of 2010 out of the sample for the purposes of comparison
within predictions that are obtained . We then found the best model
for prediction upon MSE criterion. and it was conduced that the vector
ARIMA(1,0,0)*(0,1,1)12 was the best to represent the data upon SIC
criteria. We used package software SCA which is specialized in the field of
time series to obtain the results of the application.

In this research ,the estimation of the parameters of the truncated
regression model was perfomed using two algorithms, the first algorithm is
one of the intelligent techniques from which we seek to get the optimal
solution, which is known as particle swarm optimization algorithm, and the
second is one of the conventional optimization algorithms; which is known as
Quasi-Newton algorithm namely BFGS algorithm to reach the optimum
values for these parameters. This research also proposes a hybrid algorithm,
linking BFGS algorithm with PSO algorithm. To find the optimal values for
these parameters, we are programming these algorithms using the ready
matlab7.11(R2010b). Results show that the number of iterations resulting
from the use of the hybrid algorithm (BFGS-PSO) is less than the number of
iterations of the algorithm (BFGS) and that the results were obtained using the
program Stata11 by the same amount of allowable errors.

This paper studies the time series for a number of pneumonia patient
as a Markov chain by proposing hypotheses concerning the number of
infections in order to set the problem according to Markov Chain depending
on the number of cases that are represented. After we found the statistical
characteristics for the chain that proved to be (Ergodic), then we found the
stationary distribution for it.

This paper investigates studied the reality of the total monthly demand
of Nineveh Province electricity for the period of (96) months between
(2004-2012). The statistical analyses of the studied data showed that there
is seasonal effective, general trend and random pattern .The Unobserved
Components Model (UCM) was used in the series analysis and
forecasting for four years’ prediction till 2016 was adopted.
Depending on the statistics Goodness of fit using forecasting errors,
fortunately it turn out to be a good fitting for the data model used in the
research.

One of the diagnostic methods for non-stationary dynamic systems was used in
this research , namely data segmentation. The identification problem for non-stationary
systems starts from the outline diagnostic, which considered as the cornerstone of
reaching a model that is more appropriate in describing the system . Consequently the
data segmentation operation is performed, were the input and output series are
segmented into time intervals so that stable data can be obtained within each interval.
The main objective of the data segmentation process is to minimize the loss function
and obtain a model that represents the data interval in question as best as possible. We
used the genetic algorithm to find the number, location change points, and model fits for
each part by applying a generator. Results showed that it is a good and efficient way to
estimate the location and number of points of change.

This study investigates the prediction of unstable random spatial
process using the Ordinary Kriging technique, based on the variogram
function Y(X,Y) in finding out the predictors, and the universal
multivariate technique, which has been applied by computing the Cross
Variogram Function. Practically, the data were taken from a geological
study conducted in the Department of Geography, Faculty of Education,
Mosul University. The study includes the height levels of (40) oil wells in
Kirkuk. We discovered out the estimation of the Variogram and Cross the
Variogram functions as well as the Mean Squared Error (MSE). Most of
the results have been presented in tables and figures and the results were
generally good, acceptable and close to reality,Our results showed that the
overall Kriging technique or technique that was computed by the Cross
Variogram Function was better than the Ordinary Kriging technique or
technique based on the variogram function.