Use of two-layer genetic programming for multidimensional symbolic regression

Abstract

This paper focuses on exploring the potential benefits and advantages or disadvantages of a two-layer approach in genetic programming. The first section describes two-layer genetic programming itself and how it differs from its basic version. The Python programming language framework DEAP was used for the implementation. The focus of the paper is also to compare the results obtained by using this two-layer genetic programming with different configurations of the parameters with ordinary basic genetic programming on different multidimensional datasets and benchmarks.