Statistics is a weird special case where major subfields of applied statistics (including machine learning, but not only) sometimes retain wildly divergent terminology for the exact same concepts, for no good reason at all except the vagaries of historical development.