A novel autonomous evolutionary algorithm for construction of decision trees is presented together with an analysis of different medic al datasets. The algorithm's capability to self- adapt to a given problem is used as a measure to predict if some dataset is just difficult or impossible to analyze. If a specific dataset doesn't include enough or proper information for a creation of a good general decision model then the overfitting will occur. To detect overfitting we can use several existing techniques, the most common uses special testing data that is excluded from the learning phase. The autonomous algorithm on average produces very general solutions or gives no solution if the dataset is prone to overfitting.