Towards D-Optimal Input Design for Finite-Sample System Identification

Kolumbán, Sándor and Csáji, Balázs Csanád (2018) Towards D-Optimal Input Design for Finite-Sample System Identification. In: 18th IFAC Symposium on System Identification, July 9-11, 2018, Stockholm, Sweden.

FSID-Input-Design-SYSID-2018.pdf - Published Version

Download (323kB) | Preview


Finite-sample system identification methods provide statistical inference, typically in the form of confidence regions, with rigorous non-asymptotic guarantees under minimal distributional assumptions. Data Perturbation (DP) methods constitute an important class of such algorithms, which includes, for example, Sign-Perturbed Sums (SPS) as a special case. Here we study a natural input design problem for DP methods in linear regression models, where we want to select the regressors in a way that the expected volume of the resulting confidence regions are minimized. We suggest a general approach to this problem and analyze it for the fundamental building blocks of all DP confidence regions, namely, for ellipsoids having confidence probability exactly 1/2. We also present experiments supporting that minimizing the expected volumes of such ellipsoids significantly reduces the average sizes of the constructed DP confidence regions.

Item Type: Conference or Workshop Item (Paper)
Subjects: Q Science / természettudomány > QA Mathematics / matematika > QA75 Electronic computers. Computer science / számítástechnika, számítógéptudomány
T Technology / alkalmazott, műszaki tudományok > TK Electrical engineering. Electronics Nuclear engineering / elektrotechnika, elektronika, atomtechnika
Depositing User: Dr. Balázs Csanád Csáji
Date Deposited: 05 Oct 2018 07:12
Last Modified: 05 Oct 2018 07:12

Actions (login required)

Edit Item Edit Item