The Maximum of the Periodogram of a Non-Gaussian Sequence

作者: Richard A. Davis , Thomas Mikosch

DOI: 10.1214/AOP/1022677270

关键词:

摘要: It is a well-known fact that the periodogram ordinates of an lid mean-zero Gaussian sequence at Fourier frequencies constitute exponential vector, hence maximum these has limiting Gumbel distribution. We show for non-Gaussian mean-zero, finite variance this statement remains valid. also prove point process constructed from converges to Poisson process. This implies joint weak convergence upper order statistics ordinates. These results are in agreement with empirically observed phenomenon various functionals have very much same asymptotic behavior as applied sample.

参考文章(17)
R.-D. Reiss, Approximate Distributions of Order Statistics Springer Series in Statistics. ,(1989) , 10.1007/978-1-4613-9620-8
J. Franke, W. Hardle, On Bootstrapping Kernel Spectral Estimates Annals of Statistics. ,vol. 20, pp. 121- 145 ,(1992) , 10.1214/AOS/1176348515
Hong-Zhi An, Zhao-Guo Chen, E.J Hannan, The maximum of the periodogram Journal of Multivariate Analysis. ,vol. 13, pp. 383- 400 ,(1983) , 10.1016/0047-259X(83)90017-9
David Freedman, David Lane, The empirical distribution of the fourier coefficients of a sequence of independent, identically distributed long-tailed random variables Zeitschrift f�r Wahrscheinlichkeitstheorie und Verwandte Gebiete. ,vol. 58, pp. 21- 39 ,(1981) , 10.1007/BF00536193
D. Janas, R. Dahlhaus, A frequency domain bootstrap for ratio statistics in time series analysis Annals of Statistics. ,vol. 24, pp. 1934- 1963 ,(1996) , 10.1214/AOS/1069362304
E.J. Hannan, M. Mackisack, A law of the iterated logarithm for an estimate of frequency Stochastic Processes and their Applications. ,vol. 22, pp. 103- 109 ,(1986) , 10.1016/0304-4149(86)90117-1
David Freedman, David lane, The Empirical Distribution of Fourier Coefficients Annals of Statistics. ,vol. 8, pp. 1244- 1251 ,(1980) , 10.1214/AOS/1176345197
Uwe Einmahl, Extensions of results of Komlo´s, Major, and Tusna´dy to the multivariate case Journal of Multivariate Analysis. ,vol. 28, pp. 20- 68 ,(1989) , 10.1016/0047-259X(89)90097-3