Viterbi

Implements the Viterbi algorithm[#!rabiner!#]. Given an observation sample, determines the sequence of hidden states, known as state path, that is most likely to have generated it.

Usage: viterbi( $<$object_1$>$ ,$<$object_2$>$ )

where $<$object_1$>$ is the MTK object containing the observation sample; and $<$object_2$>$ is the MTK object used to store the state path evaluated.

Output: Prints the initial and the last state of the state sequence determined, and the log-likelihood (under the name Score) that the observation sample was generated by that particular sequence of states. As mentioned above, the complete state path is stored in $<$object_2$>$.