Strong consistent and asymptotically normal estimate of parameter of first order autoregression process with infinite variance
One considers stationary first orderautoregression process and proposes strong consistent estimate its parameter, that doesnt demand existence of moments of distributionfunction of initial process. It was shown, that for asymptotic normality of this estimate only existence of first moment of initialprocess is nessesary.
Download file
Counter downloads: 296
Keywords
Authors
| Name | Organization | |
| Kitayeva A.V. | Tomsk State University | olz@mail.tomsknet.ru |
| Terpugov A.F. | Tomsk State University | terpugov@fpmk.tsu.ru и terpugov@ic.tsu.ru |
References
Samorodnitsky G., Taqqu M.S. Stable Non-Gaussian Random Processes: Stochastic Models with Infinite Variance. London: Chapmen and Hall, 1997.
Peters E.E. Fractal Market Analysis. Wiley & Sons, 1994.
Bloomfield P., Steiger W. Least Absolute Deviations: Theory, Applications and Algorithms. Boston: Birkhouser, 1983.
Хеннекен П.Л., Тортра А. Теория вероятностей и некоторые ее приложения. М.: Наука, 1974.
Королюк В.С., Портенко Н.И. и др. Справочник по теории вероятностей и математической статистике. М.: Наука, 1985.
