liveBook · Manning

Hidden Markov model

As explained in Timeless Algorithms, The Foundational Ideas

A hidden Markov model is a probabilistic model of sequences in which hidden states evolve according to transition probabilities and generate observable outputs.

A hidden Markov model (HMM) represents sequential data whose underlying states are not directly observed. Hidden states evolve through transition probabilities and generate observable outputs. section 3.6.3

Estimating its parameters is challenging because the hidden variables prevent a closed-form solution. The Baum–Welch algorithm applies expectation-maximization (EM), iteratively increasing the likelihood of the observed sequences. section 3.6.3

Applications include:

How Timeless Algorithms, The Foundational Ideas covers it

How deep it goes
Explains hidden-variable estimation through Baum–Welch and EM, but gives no implementation or step-by-step derivation.
Examples
Names speech recognition, natural language processing, biological sequence analysis, and financial modeling as application areas.
What you can do after
Understand how hidden variables complicate likelihood estimation and how Baum–Welch iteratively increases observed-data likelihood.
How it connects
Connects maximum likelihood estimation to expectation-maximization, Baum–Welch, and model selection.

Also explained in

Compare how 2 books explain hidden Markov model →

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Timeless Algorithms, The Foundational Ideas. The section links point to the parts of the book it draws on.