Complexity penalized least squares estimators: Analytical results

Abstract

Abstract A simple variational approach to the estimation of timeseries is studied in detail and mathematical rigor. The functional in question is a complexity penalized sum of squares. The results include existence and uniqueness of the statistical estimate, as well as continuous dependence and stability in dependence on parameters and data. (© 2008 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)