الفهرس 
مقدمة الدراسة والأستعراض المرجعييوجد فقط 14 صفحة متاحة للعرض العام
 |  | from | 126 |  |  |
| | | from | 126 | | |
المستخلص
The Bootstrap Method is Considered one of the Estimation Methods Based on the Principle of Sampling with Return, which is a Re-preview of a Sample of (n) Elements Drawn Randomly from (N) from the Original Data. Therefore, this Technique was Used in Linear Regression Models when there is an Autocorrelation Problem. (20, 30, 40, 50,100, 200, 500,1000) and Several Levels of the Correlation Coefficient (ρ) (0.1,…, 0.9) Depending on the Programming Language (R), The Studied Method (Bootstrap) was Compared with the Common Method to Address the Problem of Autocorrelation in Multiple Linear Regression Models (GLS). The Second Section Includes three Sxplanatory Variables and the Same Sample Sizes and Levels (ρ) for Both Sections were Studied. Based on the Comparison Criteria (CI, MSE and MAPE), the Study Found that the Bootstrap Technique is Almost Equivalent to the known Method of Addressing the Problem of Autocorrelation, Namely the Generalized Least Squares (GLS) Method with a Slight (Simple) Preference for the GLS Method Over the Studied Method. (Bootstrap) in small Samples with Fading differences Between the Two Methods with Increasing Sample size and Level (ρ) The Study also Showed a Slight Preference Method Bootstrap Method in Large Samples.