 This paper proposes an improved many objective optimization algorithm called NSGA2-RF, which combines the NSGA2 algorithm with a three-stage feature engineering process. This algorithm maximizes accuracy while simultaneously minimizing the size of the optimal solution set. Additionally, it uses multiple chromosomes hybrid coding to synchronously select features and optimize model parameters. The experimental results demonstrate that the proposed algorithm outperforms other algorithms in terms of accuracy, running time, and the size of the optimal solution set. Furthermore, the proposed algorithm also has the advantage of being more interpretable than deep learning models. This article was authored by Xiaohua Zeng, Jieping Kai, Changzhou Liang, and others.