Abstract:
|
Many data sets in biology, medicine, economic, business, fMRI and other areas deal with multiple sets of multivariate time series. A class of matrix autoregressive (MAR) models is introduced for dealing with the situation where there are multiple sets of multivariate time series data. However, the large number of the parameters is the main concern with MAR models. To overcome computational constraints, it is desired to use Bayesian approach with prior information to shrink the large number of parameters.
|