Ensembles are a set of classification models that, when combined, produce better predictions than when used by themselves. This chapter proposes
a new evolutionary algorithm-based method for creating an ensemble of rule sets consisting of two stages. First, an evolutionary algorithm (more precisely, a genetic programming algorithm) is used to automatically create complete rule
induction algorithms. Secondly, the automatically-evolved rule induction algorithms are used to produce rule sets that are then combined into an ensemble. Concerning this second stage, we investigate the effectiveness of two different
approaches for combining the votes of all rule sets in the ensemble and two different approaches for selecting which subset of evolved rule induction algorithms (out of all evolved algorithms) should be used to produce the rule sets that will be combined into an ensemble. |