Impact of the periodicity and trend on the FD parameter estimation
Publikation/Tidskrift/Serie: Journal of Statistical Computation and Simulation
Förlag: Taylor & Francis
It is well known that one of the features of long-memory processes is that they tend to have what looks like trends and cycles. A consequence of this property is that it is difficult to distinguish a long-memory process from a nonstationary process. In this paper, we study the impact of the periodicity and trend on different methods for estimating the long-memory processes parameter d.
- Medicine and Health Sciences
- Mathematics and Statistics
- discrete wavelet transform
- ISSN: 0094-9655