see the original item page
in the repository's web site and access all digital files if the item*

1991 (EN)
The eigenproblem formulation for HMM (EN)

Dologlou, I (EN)
Touratzidis, L (EN)
Carayannis, G (EN)

N/A (EN)

This paper presents a new matrix formulation of the basic concepts governing discrete Hidden Markov Models (HMM). Using this formulation, we show that symbol and state probabilities are exponential functions of the transition matrix of the model. Furthermore, based on the eigenanalysis of the transition matrix, a closed form relationship is derived between the eigenvalues of this matrix and the symbol probabilities at different instants. The matrix formulation provides a useful tool for the physical interpretation of the learning and decision process through HMM. A better insight is obtained, and tools are also given for a design with improved learning characteristics. © 1991. (EN)


Symbol/State Probabilities (EN)
speech recognition (EN)
Mathematical Techniques--Matrix Algebra (EN)
Hidden Markov model (EN)
Speech (EN)
Eigenproblem Formulation (EN)
Mathematical Techniques--Algorithms (EN)
Hidden Markov Models (HMM) (EN)
Transition Matrix (EN)

Εθνικό Μετσόβιο Πολυτεχνείο (EL)
National Technical University of Athens (EN)

Speech Communication (EN)



*Institutions are responsible for keeping their URLs functional (digital file, item page in repository site)