Tensor Factor Model Estimation by Iterative Projection

Tensor time series, which is a time series consisting of tensorial observations, has become ubiquitous. It typically exhibits high dimensionality. One approach for dimension reduction is to use a factor model structure, in a form similar to Tucker tensor decomposition, except that the time dimension is treated as a dynamic process with a time dependent structure. In this paper we introduce two approaches to estimate such a tensor factor model by using iterative orthogonal projections of the original tensor time series. The approaches extend the existing estimation procedures and our theoretical investigation shows that they improve the estimation accuracy and convergence rate significantly. The developed approaches are similar to higher order orthogonal projection methods for tensor decomposition, but with significant differences and theoretical properties. Simulation study is conducted to further illustrate the statistical properties of these estimators.