Summary: | This is a report on the results of our experiments with hidden Markov models, focusing on Markov chains with generative transitions and the Baum-Welch algorithm. We explore generating the hypothesis model in various ways. We use the hypothesis model as the original model. And we investigate the feasibility of implementing a sequential version of the Baum-Welch algorithm. This research was motivated by the ongoing research on Active Learning of Markov Decision Processes using Baum-Welch algorithm, at the Department of Computer Science, Reykjavík University.
|