Autoregressive HMM implementation?

Hello MATLAB community,

Is there any Autoregressive hidden Markov model (ARHMM) implementations available in MATLAB? I know that there are AR model functions but I cannot find any for the HMM.

Ashley

Answer 1

[Aman]

Hi @Ashley Conard,

I didn't find any out-of-box implementation available for ARHMM in MATLAB, but in order to implement it in MATLAB, you can follow the below steps:

1. Define the number of hidden states, the order of the AR process, and initialize the transition matrix, emission parameters (AR coefficients), and state probabilities.
2. Use the Expectation-Maximization algorithm to iteratively estimate the parameters of the ARHMM.
3. Use the Forward-Backward algorithm to compute the probabilities of the hidden states given the observed data.
4. Update the AR coefficients and transition probabilities based on the results of the EM algorithm.

I hope this will help you to proceed ahead with your workflow :)