ADFGP-based robust automatic programming system-implementation and evaluation

The paper presents an approach to genetic programming (GP). The objective of the study originates in the fact that a manually produced program tends to be divided into some subroutines and reuses the subroutines frequently. In traditional GP, the program is structured by one sequence. Moreover there is no room to reuse the subroutines in traditional GP. A few techniques have been proposed for division of the program into subroutines, which attempt to discover certain subroutines. However, the reusability of GP hasn't been discussed so far. The authors propose an approach for reusability. The proposed method has a library for keeping some subroutines in order to share and reuse them. They make use of the Hanoi Tower problem to indicate the efficiency experimentally.