DHMM

Overview

This library provides a C++ implementation of Hidden Markov Models with Discrete observations. Includes MEX files for use of the HMM class under Matlab.

Features

• Class encapsulating HMM with discrete observations
• Implements the 3 key HMM algorithms: Forward algorithm, Viterbi algorithm, and Baum-Welch algorithm
• Configurable number of states
• Configurable number of observations
• Configurable topology
• Simple to use
• No dependency (except for MEX files - implementation in progress)
  
• Crossplatform (plain C/C++)
• Verified against Kevin Murphy's Matlab HMM toolbox (only discrete scalar observations)

Limitations

• Does not support vector of observations

License

Modified BSD License

History

Version	Release date	Changes
1.0	25.01.2009	Initial release

Download

Available here.