Computational issues in parameter estimation for stationary hidden Markov models.
Jan BullaAndreas BerzelPublished in: Comput. Stat. (2008)
Keyphrases
- hidden markov models
- parameter estimation
- computational issues
- maximum likelihood
- non stationary
- least squares
- model selection
- markov random field
- statistical models
- expectation maximization
- em algorithm
- speech recognition
- maximum likelihood estimation
- conditional random fields
- parameter estimation algorithm
- structure learning
- markov model
- approximate inference
- sequential data
- hidden states
- viterbi algorithm
- posterior distribution
- machine learning
- sequence classification
- continuous hidden markov models
- markov models
- gibbs sampling
- hidden state
- markov chain monte carlo
- three dimensional
- feature selection