Periodically correlated modeling by means of the periodograms asymptotic distributions (Q1685304): Difference between revisions

The authors introduce a test statistics to test whether a discrete time periodically correlated model with a given spectral density explains an observed time series. Their proposed testing procedure is based on an application of the asymptotic distribution of the periodogram established in [the second author and \textit{M. Azimmohseni}, J. Stat. Plann. Inference 137, No. 4, 1236--1242 (2007; Zbl 1107.62100)]. They make comparisons between their suggested procedure and the methods that were proposed by \textit{E. Broszkiewicz-Suwaj} et al. [``On detecting and modeling periodic correlation in financial data'', Phys. A 336, No. 1--2, 196--205 (2004; \url{doi:10.1016/j.physa.2004.01.025})]. It is observed that their testing procedure is more powerful. The performance of the proposed methods is illustrated in real and simulated data sets.