Early Warning Modelling of Employee Turnover Based on Psychological Analysis and Xgboost Algorithm
Download as PDF
Zhang Biwei, Chen Yiwen
Employee’s turnover behavior contrary to the enterprise's will often means great losses to the enterprise. It is of great strategic significance for enterprises to reduce employee turnover as much as possible at the least cost. This paper proposes a new employee turnover early warning model and application process. Firstly, the factors related to employee turnover are preliminarily selected based on psychological analysis. Secondly, the attributes related to employee resignation are further selected by data preprocessing and Chi-square test. Then, the Xgboost algorithm is used to build an employee turnover early warning model. Finally, the early warning model is tested in the actual scenario of an Internet enterprise. The test results show that the effectiveness of the early warning model is 92.5%, which is recommended for large-scale application in the actual landing work.
Psychological analysis, Employee turnover, Early warning model, Xgboost