Hidden Markov Model (HMM)

overview

Summary

A Hidden Markov Model (HMM) is a probabilistic model for sequences where the underlying states are hidden but generate observable outputs. It assumes a first order Markov process over hidden states and that each observation depends only on the current state. Core tasks include scoring a sequence, decoding the most likely state path, and learning parameters from data. HMMs are classic, interpretable tools used in speech, NLP, bioinformatics, and time series.