江汉大学学报（自然科学版）

决策树分类是数据挖掘中的一种重要方法。探讨了决策树算法的基本思想和常用算法,并将决策树挖掘技术应用于大学生心理健康数据,分析挖掘影响大学生心理健康的因素。文章选择C5.0算法,通过Clementine12.0进行决策树挖掘模型的构建,建立数据流,通过不断测试分析,发现影响大学生心理健康主要症状是强迫症。以强迫症为分类目标查看模型,可以了解到焦虑症和人际关系也起到很大的影响作用。将目标属性分别设置为焦虑_程度和人际关系_程度,输出变量设为剩余的9个因子变量,执行数据流挖掘出导致强迫症的主要原因,为指导心理健康的工作人员提供参考。

Classification of decision tree is an important method in data mining. The basic ideas andcommon algorithms of decision tree algorithm are discussed,the decision tree mining is applied tostudents′ mental health data analysis,and to analyse the impacting factors on students′ mental health.With the C5.0 algorithm, performed by Clementine 12.0, the decision tree mining model wasconstructed,the data flow was also set,with continuous test and analysis,discovered that compulsionwas the main symptom which impacted the mental health of students. To view the model with compulsionas the classification object,it can be find out that anxiety and social relationship also have big influences.The target attribute were set as anxiety_degree and social relationship_degree,output variables were setas the left nine factors,dug out the main factors which cause the compulsion,to provide the reference tothe mental health domain.