Testing for the presence of self-similarity of Gaussian series having stationary increments (Q2742777)

The author proposes a procedure of testing for the presence of self-similarity of a Gaussian time series with stationary increments. The test is based on estimation of the distance between the time series and a set of time series containing all the fractional Brownian motions. This distance \(CM^2(N)\) is constructed from two estimators of multiscale generalized quadratic variations expectations. The second one requires regression estimates of the self-similarity index H.NEWLINENEWLINENEWLINEThus, two estimations of H are introduced (ordinary least squares estimates, OLSE, and generalized least squares GLSE). Both LSEs present good robustness and computing time properties compared with the Whittle contrast estimator, with nearly similar convergence rates. It is also proved that under self-similarity assumptions the distribution of the test statistic \(CM^2(N)\) converges to a \(\chi^2\) distribution and so provides a simple testing procedure. Finally, the proposed method is applied to the medical examples that motivated this research.