Joint CIAM and Optimization and Systems Theory Seminar
Friday May 9, 2008, 11.00-12.00, Room 3721, Lindstedtsvägen 25

Augusto Ferrante, Universita di Padova

Multivariate spectrum approximation in the Hellinger distance

We present a matricial version of a generalized moment problem with degree constraint. We introduce a new metric on multivariable spectral densities induced by the family of their spectral factors which, in the scalar case, reduces to the Hellinger distance. We solve the corresponding constrained optimization problem via duality theory. A highly nontrivial existence theorem for the dual problem is established in the Byrnes-Lindquist spirit. A globally convergent  matricial Newton-type algorithm is provided for the numerical solution of the dual problem. This procedure is applied to multivariate spectral estimation. Simulations show that, in the case of short observation records, this method may provide a valid alternative to standard multivariable identification techniques such as MATLAB's PEM and MATLAB's N4SID.

Calendar of seminars Last update: April 21, 2008 by Marie Lundin.