Abstract
We give novel conditions for the existence of the limit of the normalized log-likelihood function for a finite-state continuous read-out Hidden Markov Model. Our results complements those of [8]. The result is then applied to derive a strong approximation result for the parameter estimates of the Hidden Markov Model.
This is a preview of subscription content, log in via an institution to check access.
Preview
Unable to display preview. Download preview PDF.
Author information
Authors and Affiliations
Editor information
Rights and permissions
About this paper
Cite this paper
Gerencsér, L., Molnár-Sáska, G. Estimation and Strong Approximation of Hidden Markov Models. In: Benvenuti, L., De Santis, A., Farina, L. (eds) Positive Systems. Lecture Notes in Control and Information Science, vol 294. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-44928-7_42
Download citation
DOI: https://doi.org/10.1007/978-3-540-44928-7_42
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-40342-5
Online ISBN: 978-3-540-44928-7
eBook Packages: Springer Book Archive