An algebraic analysis using Matrix Padé Approximation to improve the choice of certain parameter in Scalar Component Models

This paper presents an algebraic analysis using Matrix Padé Aproximation to improve the identification stage of the proposal in [6] on Scalar Component Models, specifically as it refers to the choice of a parameter they denote h. The original methodology in [6] is based on the construction and interpretation of a table whose elements are related to the singular value zero of certain relevant matrices in the process. We propose the alternative use of what we call a Ranks Table and the sure overall orders concept instead of the so-called overall orders. Ranks Table information allows for the improved interpretation and implication of the results and of potential computational and statistical properties.