Fits latent (hidden) Markov models on mixed categorical and continuous (time series) data, otherwise known as dependent mixture models.
