Assembly: Accord.Statistics (in Accord.Statistics.dll) Version: 188.8.131.52 (184.108.40.20632)
The Viterbi learning algorithm is an alternate learning algorithms for hidden Markov models. It works by obtaining the Viterbi path for the set of training observation sequences and then computing the maximum likelihood estimates for the model parameters. Those operations are repeated iteratively until model convergence.
The Viterbi learning algorithm is also known as the Segmental K-Means algorithm.ViterbiLearning TDistribution BaumWelchLearning