A differential evolutionary chromosomal gene expression programming
Paper Url: https://www.sciencedirect.com/science/article/pii/S1568494623001114
This study addresses the need for consistent performance in automated data-driven modeling tools for intelligent system applications. To overcome limitations in Classical Gene Expression Programmings (GEPs), a Differential Evolutionary Chromosomal GEP (DEC-GEP) algorithm is proposed. DEC-GEP utilizes the Differential Evolution (DE) algorithm to optimize both expression tree genotypes and numerical values of constant terminals simultaneously, enhancing search efficiency. The modified chromosome structure includes a modifier gene container for storing constant terminal values. The DEC-GEP algorithm demonstrates its consistency in a data-driven modeling application, particularly in soft calibration of low-cost solid-state sensor array measurements. Experimental results show that DEC-GEP provides reliable CO concentration estimation models for electronic nose applications.
