A rationale for model selection criteria ─ Informational Complexity (ICOMP) Criteria, that combine a badness-of-fit term with a measure of complexity of a model, is introduced. The ICOMP criterion suggested by Bozdogan is seen as an approximation to the sum of two Kullback-Leibler distances. The asymptotic consistency properties of the class of ICOMP criteria are investigated first in the case when one of the models considered is the true model, and then in the case when none of the models is the true model. With finite sample size, the model selected by ICOMP is closer to the true model than the ones obtained by existing methods.
LV Chunlian. , {{custom_author.name_en}}.
On Informational Complexity Criteria. Journal of Systems Science and Mathematical Sciences, 2008, 28(6): 758-768 https://doi.org/10.12341/jssms10211
PDF(440 KB)
