Autonomous evolutionary algorithm in medical data analysis

Abstract

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