Author's: T. J. O’neill, Jack Penm and R. D. Terrell
Pages: [11] - [25]
Received Date: November 24, 2010
Submitted by:
An innovative set of recursive algorithms for two-dimensional (2D)
subset autoregressive (AR) models, including full order 2D ARs, is
explored and developed by progressing further from the 2D full order
AR recursions. After the initialisation is carried out by the direct
method, the proposed recursions then allow us to develop successively
all 2D subset ARs, of size (the number of lags with nonzero coefficient
matrices) from 1 to K. Finally, the best subsets of each size
with the minimum generalised residual power for that size are compared
by employing a model selection criterion to find the optimum 2D subset
AR.
recursions, subset autoregressions, two-dimensional recursive algorithms.