Gene Expression Programming: A New Adaptive Algorithm for Solving Problems

Title: Gene Expression Programming: A New Adaptive Algorithm for Solving Problems

Research Question: Can Gene Expression Programming (GEP) be a more efficient and adaptable method for solving a variety of problems compared to existing techniques?

Methodology: The study introduces GEP, a new genetic algorithmic technique. It uses linear chromosomes composed of genes structurally organized in a head and a tail. These chromosomes function as a genome and are subjected to modification through mutation, transposition, root transposition, gene transposition, gene recombination, one-point, and two-point recombination. The chromosomes encode expression trees, which are the objects of selection. The separation of genome and expression tree functions allows the algorithm to operate efficiently and adaptably.

Results: The study demonstrates GEP's effectiveness by applying it to a range of problems, including symbolic regression, sequence induction with and without constant creation, block stacking, cellular automata rules for the density-classification problem, and two problems of Boolean concept learning: the 11-multiplexer and the GP rule problem. The results show that GEP performs well and is versatile in solving these problems.

Implications: The results suggest that GEP is a promising technique for problem-solving due to its adaptability and efficiency. It surpasses existing adaptive techniques and has the potential to move beyond the phenotype threshold, a crucial stage for life's evolution. The study also highlights the importance of maintaining a functional genotype/phenotype relationship in artificial life systems.

Link to Article: https://arxiv.org/abs/0102027v2 Authors: arXiv ID: 0102027v2