 This study compared two algorithms for identifying gate events before they occur. One algorithm used a heuristic approach, which identified features related to initial contact and toe-off with high accuracy. The other algorithm used a beta process autoregressive hidden Markov model, BPARHMM, which identified features related to both initial contact and toe-off with even higher accuracy. Both algorithms had similar levels of accuracy and could potentially be used for classification and prediction of locomotion mode. This article was authored by Seth Ardonahue and Michael E. Hahn.