本科毕业生主动离职的调查实证——基于决策树模型的研究-《集美大学学报（教育科学版）》

文章信息/Info

Title:: Decision Tree modelbased Investigation of Voluntary Turnover among Undergraduates

Author(s):: ZHANG Xue -xin; (School of Mathematics and Statistics, Hubei Engineering University, Xiaogan 432000, China)

Keywords:: undergraduate; voluntary turnover; decision tree model; information gain rate; confidence; lift curve

摘要:: 用2014年应届本科毕业生的跟踪调查数据，挖掘大学毕业生主动离职的主因，建立决策树模型预测本科毕业生主动离职倾向。结论是：“毕业后三个月的薪酬”、“就业观”、“对高校设置就业指导课的作用性评价”是影响毕业生主动离职的三大主要因素。一条重要的推理规则是：如果“毕业后三个月的薪酬在2000元以下或者2001至3000元”，则“主动离职”。该决策树模型与神经网络模型、逻辑斯蒂模型对比评价总体良好，可以用于预测大学本科毕业生的主动离职倾向，为企业选定人才及毕业生就业正确定位提供参考。

Abstract:: This paper mined the important causes of graduate’ voluntary turnover by application of a tracking surveys data offered by Admission and Employment Office from the fresh graduates in 2014, using a decision tree model to predict undergraduate’s tendencies to leave their jobs. The conclusion is: The wage level after three months from graduation, The view of employment and the appraise of role of employment guidance course set up in college are three main factors affecting undergraduates’ voluntary demission. An important association rule is that if a student’s wage level after three months from graduation is below RMB 2000 Yuan per month or RMB 2001-3000 per month, He/she will voluntary leave job. This decision tree model behaves well in comparison with Artificial Neural Network and Logistic model，So it can be used to predict undergraduates’ a higher propensity to leave their current jobs for enterprise selecting talents and undergraduates right-orientation in their choosing employment.