Relationship between Governance Structure and Performance Based on the Application of XGBoost Algorithm.

The purpose of this paper is to investigate the relationship between the governance structure and performance of commercial banks, and to optimize the governance structure of listed commercial banks and thus promote the overall performance of listed commercial banks. In this paper, the return on net assets of listed commercial banks in China's Shanghai and Shenzhen stock markets over the period 2016 - 2020 is the dependent variable, and the proportion of independent directors, leadership structure, Z-index, number of board members and the percentage of largest shareholders are independent variables required for the empirical model to conduct data analysis. According to the literature review, the linear regression relationship of the model applies to firms, but whether it is applicable in commercial banks. After applied linear regression and finds that the linear relationship is not applicable in commercial banks. Further it develops a model by using the eXtreme Gradient Boosting (XGBoost) algorithm to illustrate the relationship between the governance structure and performance of commercial banks. Finally, The paper will discuss and proposes suggestions based on the findings, intending to optimize the governance structure of listed commercial banks and thereby promote the overall performance management standards of listed commercial banks. In this paper, it found that the relationship between governance structure and performance in commercial banks is not as linear as other corporations, and a more robust relationship model is obtained by using the XGBoost algorithm. After using the XGBoost algorithm, the analysis results are better. The root means the square error is only 0.003 and the fit of the regression model is over 99%, which is significantly better than traditional linear regression analysis.