This paper identifies U.S. monetary and fiscal dominance regimes using machine learning techniques. The algorithms are trained and verified by employing simulated data from Markov-switching DSGE models, before they classify regimes from 1968-2017 using actual U.S. data. All machine learning...