首页 | 本学科首页   官方微博 | 高级检索  
检索        

决策树和回归技术在中小学生视力影响因素分析中的应用
引用本文:谢小莲,李娟,刘尚红,孙丽姣,赵海萍.决策树和回归技术在中小学生视力影响因素分析中的应用[J].中国学校卫生,2019,40(4):572-575.
作者姓名:谢小莲  李娟  刘尚红  孙丽姣  赵海萍
作者单位:宁夏医科大学公共卫生与管理学院,银川,750004
摘    要:目的采用决策树模型和Logistic回归模型了解宁夏中小学生视力状况及其影响因素,为制订视力不良防控方案提供依据。方法采用分层整群抽样的方法,从宁夏银川和吴忠分层整群随机抽取7~18岁儿童3 257名作为研究对象,对学生进行视力检查和问卷调查。应用Clementine 12.0中的C 5.0算法和Logistic回归分析视力影响因素,利用灵敏度、特异度及正确分类百分比评价2个模型的效果。结果宁夏中小学生视力不良总体检出率为62.6%,男生、女生视力不良检出率分别为56.8%,68.1%,差异有统计学意义(P<0.01);不同学段学生视力不良检出率差异有统计学意义(P<0.01),其中高中生检出率最高,达82.2%;决策树C 5.0算法分析结果显示,学段分组、性别、每日做作业的时间及每周进行中等强度锻炼的频率是中小学生视力不良最主要的影响因素;Logistic回归结果显示,性别女(OR=1.65,95%CI=1.40~1.96)、眼睛距计算机显示屏距离<66 cm(OR=1.24,95%CI=1.00~1.52)、读书时眼睛距离课本不足1尺(OR=2.05,95%CI=1.23~3.40)与视力不良呈正相关;低学段组和放学后做作业时间<2 h/d(OR=0.76,95%CI=0.62~0.92)与视力不良呈负相关。结论学段、性别和放学后作业时间是影响宁夏中小学生视力状况的主要因素。建议近视预防从低学段组入手,尽可能减少中小学生视近和课后作业时间,适量增加尤其是女生的户外活动时间,从而预防近视的发生和发展。

关 键 词:决策树  回归分析  视力    患病率  学生

Application of decision tree and regression technique in the analysis of visual acuity influencing factors in primary and middle school students
XIE Xiaolian,LI Juan,LIU Shanghong,SUN Lijiao,ZHAO Haiping.Application of decision tree and regression technique in the analysis of visual acuity influencing factors in primary and middle school students[J].Chinese Journal of School Health,2019,40(4):572-575.
Authors:XIE Xiaolian  LI Juan  LIU Shanghong  SUN Lijiao  ZHAO Haiping
Institution:(School of Public Health and Management,Ningxia Medical University, Yinchuan (750004), China)
Abstract:Objective To investigate the visual acuity of primary and middle school students in Ningxia by using decision tree model and logistic regression, and to provide evidence for the prevention and control of visual acuity. Methods Using stratified cluster sampling method, 3 257 children aged 7-18 years were randomly selected from Yinchuan and Wuzhong in Ningxia. Visual acuity test and questionnaire survey were conducted. The C 5.0 algorithm in Clementine12.0 and Logistic regression were used to analyze the factors associated with visual acuity. Results The overall rate of low vision of the middle school of Ningxia was 62.6%, with girls (68.1%) higher than that of boys (56.8%) (P<0.01). There was a significant difference in the prevalence of low vision among different educational stages (P<0.01), with high school students had the highest detection rate (82.2%). Results of decision tree C 5.0 algorithm showed that educational stage, gender, time to do homework after school and frequency of moderate intensity exercise per week associated with poor eyesight of primary and secondary school students. Logistic regression results showed that girls, the distance between the eyes and the computer display screen <66 cm (OR=1.24, 95%CI=1.00-1.52), and the distance between the eyes and the textbook when reading was less than one foot (OR=2.05, 95%CI=1.23-3.40) were positively associated with poor vision. Low grade students and homework time <2 hours after school (OR=0.76, 95%CI=0.62-0.92) were negatively associated with lower risk of poor vision. Conclusion Educational stage, gender and time of homework after school are associated with the visual acuity of primary and secondary school students in Ningxia. It is suggested that prevention of myopia should start from the low-grade students, reducing after-school work for primary and secondary school students, and increasing outdoor activities. All the girls should be encouraged to engage in more outdoor activities, to prevent the occurrence and development of myopia.
Keywords:Decision trees  Regression analysis  Vision  low  Prevalence  Students
本文献已被 维普 等数据库收录!
点击此处可从《中国学校卫生》浏览原始摘要信息
点击此处可从《中国学校卫生》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号