Article
Keywords:
RCA; parameter estimation; rate of convergence
Summary:
This work deals with Random Coefficient Autoregressive models where the error process is a martingale difference sequence. A class of estimators of unknown parameter is employed. This class was originally proposed by Schick and it covers both least squares estimator and maximum likelihood estimator for instance. Asymptotic behavior of such estimators is explored, especially the rate of convergence to normal distribution is established.
References:
[1] Basu A. K., Roy S. Sen:
On rates of convergence in the central limit theorem for parameter estimation in general autoregressive model. Statistics 21 (1990), 461–470
DOI 10.1080/02331889008802256 |
MR 1062852
[2] Basu A. K., Roy S. Sen:
On rates of convergence in the central limit theorem for parameter estimation in random coefficient autoregressive models. J. Indian Statist. Assoc. 26 (1988), 19–25
MR 1002100
[3] Davidson J.:
Stochastic Limit Theory. (Advanced Texts in Econometrics.) Oxford University Press, New York, Reprinted 2002
MR 1430804 |
Zbl 0904.60002
[6] Kato Y.:
Rates of convergence in central limit theorem for martingale differences. Bull. Math. Statist 18 (1979), 1–8
MR 0517156
[7] Michel R., Pfanzagl J.:
The accuracy of the normal approximation for minimum contrast estimate. Z. Wahrsch. Verw. Gebiete 18 (1971), 73–84
DOI 10.1007/BF00538488 |
MR 0288897
[12] Vaněček P.: Estimators of generalized RCA models. In: Proc. WDS’04 Part I (2004), pp. 35–40