Train regression, classification, tree-based, kernel, and ensemble models, choosing methods that fit the data and operational constraints.