--I have experience a working model on decision trees, predicting classification churn-no churn for a telco
--worth to say, random forests most likely can do better than decision trees alone, since random forests are compositions of decision trees, bypassing the handicaps of the latter