Efficient training of high-order hidden Markov models using first-order representations.
Johan A. du PreezPublished in: Comput. Speech Lang. (1998)
Keyphrases
- hidden markov models
- high order
- higher order
- markov models
- baum welch
- speech recognition
- low order
- discriminative training
- pairwise
- conditional random fields
- sequence classification
- gesture recognition
- viterbi algorithm
- markov random field
- training set
- sequential data
- hidden state
- minimum classification error
- probabilistic model
- dynamic programming
- feature vectors
- computational complexity
- hidden states
- stochastic context free grammars
- bayesian networks
- feature extraction