Building and evaluating an approximate model of simulation models using response surfaces using Bootstrap

Abstract

The evolution of the fundamental of the amount of arithmetic of the computer
(Computational Capability) provides many possibilities to solve many
problemsof complex and large, by relying on building simulation models for
acceptablesolutions, in spite of the advantages of simulation models banker, is
not alwaysthe appropriate way to analyze because of some cases of simulations
areexpensive and consumes a great time for the account and to overcome
theselimitations ,Has been in theSearch re-modeling simulation for one
sampleamong the queue model G G 1, and through the use of a type of
Metamodel isa first regression model depending on the matrix factorial
experiments and amultiple regression model depending on the matrix central be
retained for twoalternatives for a simulation model , Where the advantage of
thismodels forrepresentation of the relationship between inputs and outputs of
the system andclear to identify the factors most important and influential on the
output, whichmakes analysis more flexible , and validation And to test the
accuracy of atwo models by using (Bootstrap), Through this study it was found
that bootstrapregression model is best standards regression model in the
representation of a simulation of queues