If there’s any connection that I thought wasn’t mentioned in the article, is was the relationship between kalman filters and hidden markov models.