Model Hidden Markov (HMM) is a Markov statistical in which the system has been modeled for a process Markov modes observed for the (hidden) assumption. A model Hidden Markov can be used as the simplest network, the Bayesian dynamic to be considered.

In models ordinary Markov, etc. to directly by the Observer, is visible and therefore the probability of transfer between modes, the only parameters available. In a model of Hidden Markov, etc. mode, directly are not visible, but output, depending on the mode, is visible. Every state a probability distribution on the symbols output possible. So the sequence of symbols generated by a model of Hidden Markov information about the sequence of modes. Note that the adjectives 'hidden' to the sequence of modes that model from it, passes refers, not to the parameters of the model; even if the model parameters are accurately determined. models continue to 'hide'.

model Hidden Markov more because of their in reckoned the template, such as voice recognition and handwriting, diagnosis hint and move on. label this example as well. bioinformatics and ... are well-known.