Long memory with seasonal effects (Q5933668): Difference between revisions

The paper is a review of some results on long memory processes having hyperbolically damped oscillating covariance function. There are two parametric families of seasonal long memory models proposed by Granger. One of them (FARMA models) is obtained by filtering a white noise using the transfer function \((1-z)^d\), the other is based on aggregation of random parameters AR(1) processes. Some results concerning the influence of seasonality on limit theorems are presented. Methods of simulation of these models and parameter estimation are also discussed.