State-by-state Minimax Adaptive Estimation for Nonparametric Hidden Markov Models

In this paper, we introduce a new estimator for the emission densities of a nonparametric hidden Markov model. It is adaptive and minimax with respect to each state's regularity--as opposed to globally minimax estimators, which adapt to the worst regularity among the emission densities. Our method is based on Goldenshluger and Lepski's methodology. It is computationally efficient and only requires a family of preliminary estimators, without any restriction on the type of estimators considered. We present two such estimators that allow to reach minimax rates up to a logarithmic term: a spectral estimator and a least squares estimator. We show how to calibrate it in practice and assess its performance on simulations and on real data.

[1]  P. Massart,et al.  Minimal Penalties for Gaussian Model Selection , 2007 .

[2]  Jean-Marc Robin,et al.  Estimating Multivariate Latent-Structure Models , 2016, 1603.09141.

[3]  H. Weimerskirch,et al.  Movement models provide insights into variation in the foraging effort of central place foragers , 2014 .

[4]  Jean-Marc Robin,et al.  Non‐parametric estimation of finite mixtures from repeated measurements , 2016 .

[5]  A. V. Goldenshluger,et al.  General Selection Rule from a Family of Linear Estimators , 2013 .

[6]  Anima Anandkumar,et al.  A Method of Moments for Mixture Models and Hidden Markov Models , 2012, COLT.

[7]  Dean Alderucci.  A SPECTRAL ALGORITHM FOR LEARNING HIDDEN MARKOV MODELS THAT HAVE SILENT STATES , 2015 .

[8]  Yohann de Castro,et al.  Minimax Adaptive Estimation of Nonparametric Hidden Markov Models , 2015, J. Mach. Learn. Res..

[9]  Yohann De Castro,et al.  Consistent Estimation of the Filtering and Marginal Smoothing Distributions in Nonparametric Hidden Markov Models , 2015, IEEE Transactions on Information Theory.

[10]  Sylvain Arlot,et al.  A survey of cross-validation procedures for model selection , 2009, 0907.4728.

[11]  Elisabeth Gassiat,et al.  Finite state space non parametric Hidden Markov Models are in general identifiable , 2013, 1306.4657.

[12]  Ronan Fablet,et al.  Generalized Pareto for Pattern-Oriented Random Walk Modelling of Organisms’ Movements , 2015, PloS one.

[13]  Pascal Massart,et al.  Estimator Selection: a New Method with Applications to Kernel Density Estimation , 2016, Sankhya A.

[14]  Luc Lehéricy,et al.  Order estimation for non-parametric Hidden Markov Models , 2016 .

[15]  A. Goldenshluger,et al.  Bandwidth selection in kernel density estimation: Oracle inequalities and adaptive minimax optimality , 2010, 1009.1016.

[16]  Etienne Rivot,et al.  Identifying fishing trip behaviour and estimating fishing effort from VMS data using Bayesian Hidden Markov Models , 2010 .

[17]  J. Rousseau,et al.  Efficient semiparametric estimation and model selection for multidimensional mixtures , 2016, 1607.05430.

[18]  Bertrand Michel,et al.  Slope heuristics: overview and implementation , 2011, Statistics and Computing.