B1838
Title: Dimension reduction in multivariate time series
Authors: S Yaser Samadi - Southern Illinois University Carbondale (United States) [presenting]
Wiranthe Herath - Drake University (United States)
Abstract: Dimensionality reduction is very important in multivariate time series analysis because the number of parameters grows rapidly with the time series dimension. There are many dimensionality reduction techniques for time series; however, the achieved dimensionality reductions by these methods are not substantial and they usually fail to extract relevant information from a complex body of data because they fail to distinguish between information that is important to the scientific goals. We introduce a new parsimonious multivariate time series model that achieves efficient estimation by linking the mean function and covariance matrix and using the minimal reducing subspace. The results of simulation studies and real data analysis that compare the performance of the proposed model with that of the existing models in the literature will be presented.
