【摘要】 [目的]使用随机森林对职业健康监护数据和人群焦虑情况进行分析,探讨数据挖掘方法的应用。[方法]收集某企业职业健康监护数据,并使用GAD-7广泛性焦虑量表进行问卷调查,然后用随机森林对职业健康监护数据以焦虑情况为结局变量进行分类。[结果]随机森林对焦虑情况的分类效果较好,焦虑高分组错分率为14.62%,焦虑低分组错分率为5.95%,袋外数据误差率估计为10.27%。[结论]将职业健康监护数据与随机森林相结合,能够为焦虑人群的早期发现、筛选和干预提供帮助,为职业健康监护数据的利用提供新思路。更多还原

【Abstract】 [Objective] To explore the association between the anxiety and other health examination indicators gotten by occupational health surveillance,the discriminant analysis classification was done with the random forest method. [Methods]The generalized anxiety disorder of workers who received occupational health surveillance were surveyed with GAD-7questionnaire. The association between anxiety disorder and health examination indicators gotten by occupational health surveillance was evaluated. The random forest method was used for discriminant analysis classification. [Results] Random forest method had a satisfactory classification effect on anxiety disorder. The error classification of anxiety group was 14.62%,and the other group was 5.95%;OOB estimate of error rate was 10.27%. [Conclusion] With random forest method,the anxiety could be earlier screened based on the occupational health surveillance data,so that the early intervention can be given.更多还原