I am trying to fit an HMM on my behavioral data. I know how to fit it on a single experimental trial. However, I have many more trials whose data is available. I am wondering if there is a way to fit an HMM on all of my data to improve learning performance.