Consistent Order Estimation and Minimal Penalties

DOI10.1109/TIT.2012.2221122MaRDI QIDQ2989433zbMATH OpenFDO

Authors Elisabeth Gassiat, Ramon van Handel

Publication date 8 June 2017

Published in IEEE Transactions on Information Theory (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1002.1280

Mathematics Subject Classification ID

Nonparametric estimation (62G05) Markov processes: estimation; hidden Markov models (62M05) Strong limit theorems (60F15)

Abstract: Consider an i.i.d. sequence of random variables whose distribution f* lies in one of a nested family of models M_q, q>=1. The smallest index q* such that M_{q*} contains f* is called the model order. We establish strong consistency of the penalized likelihood order estimator in a general setting with penalties of order eta(q) log log n, where eta(q) is a dimensional quantity. Moreover, such penalties are shown to be minimal. In contrast to previous work, an a priori upper bound on the model order is not assumed. The results rely on a sharp characterization of the pathwise fluctuations of the generalized likelihood ratio statistic under entropy assumptions on the model classes. Our results are applied to the geometrically complex problem of location mixture order estimation, which is widely used but poorly understood.

Cited in

(5)

• Strong identifiability and optimal minimax rates for finite mixture estimation
• Will the PLS criterion for order estimation work with AML and a posteriori prediction error?
• The local geometry of finite mixtures
• Optimal estimation of high-dimensional Gaussian location mixtures
• Functional Finite Mixture Regression Models

This page was built for publication: Consistent Order Estimation and Minimal Penalties

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2989433)