Simple Implementation of the Viterbi Algorithm for training Hidden Markov Models.
This is an implementation of the Viterbi Algorithm for training Hidden Markov models based on Luis Serrano's YouTube video on the subject. This repo accompanies the video found here: https://www.youtube.com/watch?v=kqSzLo9fenk
This implementation can handle prior probabilities, and any sized probability transition matrix. It cannot handle exit probabilities though.
Choose any one of the function names in the example_sets class file and use it like so in the main:
example_sets.function_name()