Analyticity, Convergence, and Convergence Rate of Recursive Maximum-Likelihood Estimation in Hidden Markov Models.
Vladislav B. TadicPublished in: IEEE Trans. Inf. Theory (2010)
Keyphrases
- convergence rate
- maximum likelihood estimation
- hidden markov models
- convergence speed
- maximum likelihood
- global convergence
- minimum classification error
- em algorithm
- step size
- learning rate
- parameter estimation
- speech recognition
- conjugate gradient
- probability distribution
- expectation maximization
- markov models
- conditional random fields
- sequential data
- gradient method
- continuous hidden markov models
- automatic speech recognition
- density function
- faster convergence rate
- baum welch
- convergence analysis
- discriminative training
- k means
- pairwise