Maximum likelihood estimation in hidden Markov models with inhomogeneous noise
Statistics Theory
2018-10-02 v2 Methodology
Statistics Theory
Abstract
We consider parameter estimation in finite hidden state space Markov models with time-dependent inhomogeneous noise, where the inhomogeneity vanishes sufficiently fast. Based on the concept of asymptotic mean stationary processes we prove that the maximum likelihood and a quasi-maximum likelihood estimator (QMLE) are strongly consistent. The computation of the QMLE ignores the inhomogeneity, hence, is much simpler and robust. The theory is motivated by an example from biophysics and applied to a Poisson- and linear Gaussian model.
Cite
@article{arxiv.1804.04034,
title = {Maximum likelihood estimation in hidden Markov models with inhomogeneous noise},
author = {Manuel Diehn and Axel Munk and Daniel Rudolf},
journal= {arXiv preprint arXiv:1804.04034},
year = {2018}
}
Comments
31 pages, 6 figures, Accepted for publication in ESAIM Probab. Stat