Impact of the Fitness Function on the Development of Mathematical Models by Symbolic Regression

Objective: This study aims to investigate the impact of the fitness function on the development of mathematical models using Symbolic Regression based on experimental data from the decolorization process of a synthetic effluent, employing a Taguchi matrix to evaluate the efficiency of this modeling process and its application to other processes. Theoretical Framework: The study is grounded in references on advanced oxidation processes, experimental design with an emphasis on the Taguchi method, and Symbolic Regression. Method: The research methodology involved modeling and simulation, with data collection conducted through photodegradation experiments on a synthetic effluent. Results and Discussion: The results demonstrated that the fitness function affects the predictive quality of the model obtained through Symbolic Regression. This impact is highlighted in the discussion section through validation experiments. Possible discrepancies and limitations of the study are also considered in this section. Research Implications: The practical and theoretical implications of this research are discussed, providing insights into how the findings can be applied or influence practices in the field of industrial process modeling and optimization. These implications are broadly applicable across various industrial sectors that involve transformation processes. Originality/Value: This study contributes to the literature with the application of new mathematical modeling techniques employed in conjunction with Experiment Design. The relevance and value of this research are evidenced by demonstrating that the use of Symbolic Regression is viable and may be superior to the Ordinary Least Squares Method.