Statistical Analysis and Prediction of Employee Turnover Propensity Based on Data Mining

Nowadays, talent competition is becoming more and more fierce, and retaining talents is the key to the survival of enterprises. In this paper, we use enterprise employee departure data to analyze the factors that lead to employee departure by using the Logistic model and XGBoost algorithm, to predict and control the tendency of employees to leave, and to make suggestions to enterprises. The results show that the prediction accuracy and fit of the XGBoost algorithm is better. In addition, this paper finds that employee turnover is not determined by a single variable, but by several factors together. The results suggest that companies should reduce employee turnover intentions in terms of satisfaction, performance evaluation, and working time arrangement.