Evolutionary Programming Guided by Analytically Generated Seeds

Evolutionary programming is the most powerful method for inducing recursive functional programs from input/output examples while taking into account efficiency and complexity constraints for the target program. However, synthesis time can be considerably high. A strategy which is complementary to the generate-and -test based approaches of evolutionary programming is inductive analytical programming where program construction is example-driven, that is, target programs are constructed as minimal generalization over the given input/output examples. Synthesis with analytical approaches is fast, but the scope of synthesizable programs is restricted. We propose to combine both approaches in such a way that the power of evolutionary programming is preserved and synthesis becomes more efficient. We use the analytical system IGOR2 to generate seeds in form of program skeletons to guide the evolutionary system ADATE when searching for target programs. In an evaluations with several examples we can show that using such seeds indeed can speed up evolutionary programming considerably.