Abstract: Under the background of the deepening of the reform of electric power system, the electric power companies in the remote areas of western China are faced with more serious personnel loss. Employee turnover prediction has attracted more and more attention in power companies. However, traditional prediction algorithms cannot effectively solve the imbalance problem of employee turnover data set in power companies. Based on this, this paper proposes a random forest algorithm based on interval variables, using the human resources data set of Qinghai Electric Power Company from 2009 to 2017 for empirical analysis, and comparing it with the prediction results of decision trees, support vector machines, and random forest algorithms. The results show that the algorithm is more suitable for solving the problem of imbalance data and has higher prediction accuracy. At the same time, the important characteristics of employee turnover are analyzed; and it can provide decision-making basis for the human resource management of related power companies.

Key words: turnover prediction, imbalanced data, random forest

摘要： 在电力体制改革全面深化的背景下,我国西部偏远地区的电力企业面临较为严重的人员流失问题。员工离职预测越来越受到电力企业关注,然而传统预测算法无法有效解决电力企业员工离职数据集的不平衡问题。基于此,本文提出一种基于区间变量的随机森林算法,采用青海省电力公司2009~2017年人力资源数据集进行实证分析,并与决策树、支持向量机、随机森林算法的预测效果进行对比。结果表明,该算法更适合解决数据不平衡问题,具有更高的预测精度;同时分析得到员工离职的重要特征,为相关电力企业人力资源管理提供决策依据。

关键词: 离职预测, 不平衡数据, 随机森林