基于决策树的大学生科研活动与毕业走向的关联研究  

Relationship between college students' scientific research and prospects after graduation based on decision tree

作　　者：胡灵敏[1] 骆力明[1] 茅潇潇 高明[2] 

机构地区：[1]首都师范大学信息工程学院,北京100048 [2]华东师范大学软件学院,上海200062

出　　处：《计算机工程与科学》2016年第6期1275-1280,共6页Computer Engineering & Science

基　　金：国家社会科学基金重大委托项目(14&ZH0036)

摘　　要：本科生的科研活动越来越得到教育研究者的重视,然而,研究者很少关注学生科研活动与毕业走向之间的关联问题。为了探究本科生参与科研情况对其未来发展的影响,从学生毕业走向切入,挖掘影响学生未来毕业去向的重要因素。以北京市某师范高校计算机相关专业的2009级的所有学生的科研情况与毕业走向作为数据主要来源,以Weka作为模型建立和评估的辅助工具,采用J48决策树分类,进行十折交叉验证,建立了基于学生科研活动的毕业走向预测模型,并通过对模型进行验证评估,证明了模型预测准确率良好。研究结果表明,学生参与竞赛的数量、专业、作为项目负责人的项目数量、参与教学类项目的数量对学生的未来走向有着重要影响。Educational researchers are paying more and more attention to scientific research activities, however, little attention is paid to the relationship between students＇ scientific research and their future direction. In order to explore this relationship, we take the careers after graduation as the pointcut, and extract the most important components that can influence the relationship. Taking computer- related majors from a normal college in Beijing as the research sample, we build a student graduation prediction model based on students＇ scientific research, and 10-fold cross-validation is used to validate the model. Evaluation results prove that the model＇s prediction accuracy is good. The results also indicate that the number of students＇ participation in competition, students＇ majors, the number of projects as a leader, and the number of students＇ participation in projects on education have important influence on student＇s future direction.

关 键 词：科研活动 决策树 毕业走向 WEKA 

分 类 号：TP312[自动化与计算机技术—计算机软件与理论]