Some Results on Algebraic and Geometric Characterization of Linear Systems Models for Time Series Analysis

Abstract

It is shown that in the multivariate case the orders p, of the AR part, and q, of the MA part, are not invariants of the time series. Thus, it is concluded that it only makes sense to define the class of ARMA(p,p)- irreducible models, where p is the biggest of the system's Kronecker indices. This class is shown not to be a differentiable manifold, but to contain one, which is a generic subset of systems with all Kronecker indices equal to p. A formula which gives the metric tensor for riemannian manifolds of linear systems as a line integral in the complex plane is introduced for deterministic and stochastic cases and some tensors are obtained with it.