arXiv subjects
Publications and source records attributed to Alexandre Rodrigues Mesquita.
This report investigates the extension of the Bayesian Subspace System Identification method proposed in our previous work to the Multiple-Input Multiple-Output (MIMO) case. We derive new equivariant priors and posterior distributions specifically suited for the MIMO framework. Numerical results utilizing the DAISY dataset are reported to validate the approach.
Model estimates obtained from traditional subspace identification methods may be subject to significant variance. This elevated variance is aggravated in the cases of large models or of a limited sample size. Common solutions to reduce the effect of variance are regularized estimators, shrinkage estimators and Bayesian estimation. In the current work we investigate the latter two solutions, which have not yet been applied to subspace identification. Our experimental results show that our proposed estimators may reduce the estimation risk up to $40\%$ of that of traditional subspace methods.