Synopses & Reviews
This work discusses closely the making of good models from observations, offering an interdisciplinary treatment of the subject. Focusing on linear dynamic systems evolving in discrete time, it examines their importance in the development of new applications in various fields, pointing out their interconnections and potential use for workers in several disciplines. Discussions cover the significant progress made in understanding the algebraic and topological structure of linear dynamic systems, work originating in the fields of systems and control engineering, the application of the scalar output case in modeling, statistically treating data arising in signal processing, the development of algorithms for on-line and real-time calculations, and recent work by statistical time-series analysts on theory and algorithms for off-line calculation. Also looks at the required asymtotic theory associated with estimation procedures.
Table of Contents
Linear Systems and Stationary Processes.
Realization and Parameterization of Linear Dynamic Systems.
The Kalman Filter.
Maximum Likelihood Estimation of Armax Systems.
Estimating the Order of a Linear System.
Calculation of the Estimates.
Approximation by Rational Transfer Functions.
References.
Index of Notation.