discreteHMC

'discreteHMC' is a Matalb toolbox to perform numerical inference over Hidden Markov Chains. Implements forward, backward, and Viterbi algorithms.
Currently works for scalar states and observations, with both stationary and non-stationary transition and observation matrices.

To Do

• ML estimation of distributional parameters
• Inference while averaging over distribution parameters
• Add support for multi-dim states and observations