From the course: Fundamentals of Dynamic Programming
Unlock this course with a free trial
Join today to access over 23,400 courses taught by industry experts.
Inferring the most probable state sequence - Python Tutorial
From the course: Fundamentals of Dynamic Programming
Inferring the most probable state sequence
- [Instructor] So, you've modeled a real world phenomena like speech recognition as a hidden Markov model. What can you do with this model? Okay, maybe you have the model and some observations. For example, you have a bunch of sound waves that you've captured. And what you want to do is calculate the most probable sequence of hidden states that resulted in those observations. In speech recognition, that's calculating the most probable sequence of syllables that produced the observed sounds. Why is this even a hard problem? Well, let's say you're looking at each chunk of sound by itself, and you predict the speaker said o and then terr, so far so good. They're talking about the animal, but as you keep going, you find yourself predicting mo and then bile something seems off. What we should have predicted as a whole is automobile. The problem was, we looked at each part in isolation, not the entire sequence. So just going…
Practice while you learn with exercise files
Download the files the instructor uses to teach the course. Follow along and learn by watching, listening and practicing.
Contents
-
-
-
-
-
-
(Locked)
What is a Hidden Markov Model?2m 21s
-
(Locked)
Modeling a Hidden Markov Model in Python3m 20s
-
(Locked)
Inferring the most probable state sequence1m 54s
-
(Locked)
Breaking down state inference into subproblems: The Viterbi algorithm3m 36s
-
(Locked)
Implementing the Viterbi algorithm in Python4m 26s
-
(Locked)
More applications of Hidden Markov Models2m 29s
-
(Locked)
Training Hidden Markov Models2m 28s
-
(Locked)
-