Advances and Applications in Statistics
Volume 18, Issue 1, Pages 57 - 72
(September 2010)
|
|
MOMENTS OF AR(2)-MODEL PARAMETER ESTIMATORS
Jan Vrbik
|
Abstract: We describe a simple technique for computing the first few moments of estimators of the
?and
?parameters of the AR(2) model. This enables us to find a good approximation to the joint distribution of the two estimators, which represents a major improvement over the Central Limit Theorem, and can be successfully used even when sample size is relatively small. Furthermore, we can also construct a transformation of the two estimators which eliminates their skewness and thus makes the corresponding approximate distribution simpler and more accurate. |
Keywords and phrases: AR(2) model, autocorrelation, parameter estimators, sampling distribution, normal approximation. |
|
Number of Downloads: 366 | Number of Views: 1065 |
|