Code and Data Repository for "Learning Hidden Markov Models with Structured Transition Dynamics"
Simin Ma, Amin Dehghanian, Gian‐Gabriel P. Garcia, Nicoleta Serban · INFORMS journal on computing · 2024
The goal of this code is to fit a Hidden Markov Model (HMM) to data with multiple linear constraints imposed on the HMM's transition and emission probability matrices via the LDKB-FISTA algorithm. Alongside our LDKB-FISTA, we have also uploaded the scripts for running Matlab-Fimincon method, as well as the traditional Baum Welch method. The code are written in Matlab.