Typically, Tree based methods are not as competitive with the best supervised learning approaches, meaning less accurate prediction, such as linear and non-linear models and their methods.