员工晋升影响因素分析及预测研究

Analysis and Prediction of Influencing Factors of Employee Promotion

摘要: 本文以Kaggle平台提供的大型跨国公司员工晋升数据集为样本，应用多种机器学习算法构建晋升预测模型，确定影响员工晋升的重要因素。研究通过特征工程优化数据，在经过SMOTE进行数据平衡后，应用Logistic回归、决策树、随机森林、XGBoost和Stacking五种算法建模，结果显示XGBoost模型预测准确率最高(94.19%)。研究还经过SHAP分析揭示了培训评估平均分、部门、总绩效表现、上一年培训次数和服务年限是影响晋升的关键前五因素。本研究为企业人力资源管理选择预测算法和优化晋升体系提供了数据支持与针对性建议。

Abstract: Based on the employee promotion dataset of large multinational companies provided by the Kaggle platform, this paper uses a variety of machine learning algorithms to construct a promotion prediction model and identify the important factors affecting employee promotion. In this study, the data is optimized by feature engineering, and the data is balanced by SMOTE, and five algorithms are used to model the data: logistic regression, decision tree, random forest, XGBoost and Stacking, and the results show that the XGBoost model has the highest prediction accuracy (94.19%). The study also revealed that the average_training_score, department, sum_metric, no_of_trainings, and length_of_service were the top five key factors influencing promotion. This study provides data support and targeted suggestions for the selection of prediction algorithms and the optimization of promotion system for enterprise human resource management.