Advances and Applications in Statistics
Volume 45, Issue 3, Pages 181 - 200
(June 2015) http://dx.doi.org/10.17654/ADASJun2015_181_200 |
|
A tv-IVAR MODEL FOR MULTIVARIATE IRREGULAR TIME SERIES
Rogério F. Porto, Oscar E. Molina and Gladys E. Salcedo
|
Abstract: In this paper, we propose a time-varying irregular vector autoregressive (tv-IVAR) model for unequally spaced non-stationary multivariate time series extending a known related model for univariate series. We expand the functional parameters of the autoregressive matrices in a wavelet basis and estimate the coefficients by least squares after truncation at a suitable resolution level. We derive simple theoretical results and run some simulations to evaluate the estimation procedure and the model behavior on finite samples. We apply the model to a set of four biological time series observed at unequal time intervals in waters of the Beagle Channel in Argentina. |
Keywords and phrases: irregularly spaced time series, locally stationary multivariate processes, vector autoregressive model, wavelet analysis. |
|
Number of Downloads: 398 | Number of Views: 1336 |
|