Asymptotic analysis of model selection criteria for general hidden Markov models

DOI10.1016/J.SPA.2020.10.006MaRDI QIDQ1994901zbMATH OpenOpenAlexFDO

Authors Shouto Yonekura, Alexandros Beskos, Sumeetpal S. Singh

Publication date 18 February 2021

Published in Stochastic Processes and their Applications (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1811.11834

zbMATH Keywords

Akaike information criteriaBayesian information criteriahidden Markov modelsmodel evidence

Mathematics Subject Classification ID

Computational methods in Markov chains (60J22) Statistical aspects of information-theoretic topics (62B10) Point estimation (62F10) Asymptotic properties of parametric estimators (62F12) Markov processes: estimation; hidden Markov models (62M05) Time series, auto-correlation, regression, etc. in statistics (GARCH) (62M10)

Abstract: The paper obtains analytical results for the asymptotic properties of Model Selection Criteria -- widely used in practice -- for a general family of hidden Markov models (HMMs), thereby substantially extending the related theory beyond typical i.i.d.-like model structures and filling in an important gap in the relevant literature. In particular, we look at the Bayesian and Akaike Information Criteria (BIC and AIC) and the model evidence. In the setting of nested classes of models, we prove that BIC and the evidence are strongly consistent for HMMs (under regularity conditions), whereas AIC is not weakly consistent. Numerical experiments support our theoretical results.

Recommendations

• Model selection for hidden Markov chains
• Asymptotics of the maximum likelihood estimator for general hidden Markov models
• Propriétés asymptotiques de l'estimateur de maximum de vraisemblance pour des modèles de Markov cachés généraux
• Asymptotical Statistics of Misspecified Hidden Markov Models
• Asymptotic normality of the maximum-likelihood estimator for general hidden Markov models
• Consistent and asymptotically normal parameter estimates for hidden Markov models

Cites work

• A new look at the statistical model identification
• A note on composite likelihood inference and model selection
• An optimal selection of regression variables
• Analogues of Khintchine, Marcinkiewicz–Zygmund and Rosenthal Inequalities for Symmetric Statistics
• Asymptotic properties of the maximum likelihood estimation in misspecified hidden Markov models
• Asymptotic properties of the maximum likelihood estimator in autoregressive models with Markov regime
• Asymptotically efficient selection of the order of the model for estimating parameters of a linear process
• Bayes Factors
• Bridging AIC and BIC: A New Criterion for Autoregression
• Can the strengths of AIC and BIC be shared? A conflict between model indentification and regression estimation
• Consistency of the maximum likelihood estimator for general hidden Markov models
• Estimating the dimension of a model
• Exponential forgetting and geometric ergodicity in hidden Markov models
• Generalised information criteria in model selection
• Hidden Markov Models and Disease Mapping
• Hidden Markov models in finance
• scientific article; zbMATH DE number 4215168 (Why is no real title available?)
• scientific article; zbMATH DE number 3723610 (Why is no real title available?)
• scientific article; zbMATH DE number 3553528 (Why is no real title available?)
• scientific article; zbMATH DE number 2106098 (Why is no real title available?)
• Inference in hidden Markov models.
• Information criteria for selecting possibly misspecified parametric models
• Learning the distribution of latent variables in paired comparison models with round-robin scheduling
• Maximum likelihood principle and model selection when the true model is unspecified
• Mean field simulation for Monte Carlo integration
• Model Selection and Model Averaging
• Nonlinear time series. Theory, methods and applications with R examples
• Online expectation maximization based algorithms for inference in hidden Markov models
• Optimal error exponents in hidden markov models order estimation
• Particle approximations of the score and observed information matrix in state space models with application to parameter estimation
• Posterior consistency for partially observed Markov models
• Regression and time series model selection in small samples
• Risk bounds for model selection via penalization
• Selection of the order of an autoregressive model by Akaike's information criterion
• Simulated likelihood inference for stochastic volatility models using continuous particle filtering
• The Application of Hidden Markov Models in Speech Recognition
• The consistency of the BIC Markov order estimator.
• The Hartman-Wintner Law of the Iterated Logarithm for Martingales
• The maximizing set of the asymptotic normalized log-likelihood for partially observed Markov chains

Cited in

(5)

• Asymmetric hidden Markov models
• Assessing the Goodness-of-Fit of Hidden Markov Models
• Online Smoothing for Diffusion Processes Observed with Noise
• Asymptotic behavior of the maximum likelihood estimator for general Markov switching models
• Determine the number of states in hidden Markov models via marginal likelihood

This page was built for publication: Asymptotic analysis of model selection criteria for general hidden Markov models

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1994901)