论文部分内容阅读
目的通过应用R软件中决策树包rpart、party、partykit对伤害发生的影响因素进行分析,比较3种方法分析结果的差别,为伤害资料分析方法提供一种新手段。方法根据农村女性伤害发生专题调查资料,在R软件中分别建立决策树包rpart、party、partykit的模型,比较3种方法分析结果的差异。结果 rpart包与party、partykit包分类规则不一致。rpart包分类结果表明未接受教育的农民女性是伤害的高危人群,party包与partykit包分类结果表明接受教育女性中离婚或丧偶的女性伤害发生率最高,其次是未接受教育的女性。结论R软件决策树包的应用条件有所不同,应该根据实际情况选择合适的软件包进行应用分析。在伤害资料的分类建模分析中,使用partykit包较为适合、简捷且界面友好。
Objective To analyze the influencing factors of injury by using decision tree package rpart, party and partykit in R software, and to compare the differences between the three methods to provide a new method for the analysis of injury data. Methods Based on the survey data of female victims in rural areas, the models of decision tree package rpart, party and partykit are respectively established in R software, and the differences between the three methods are analyzed. Results rpart package and party, partykit package classification rules are inconsistent. The results of the rpart package showed that the uneducated peasant women were the most at-risk groups. The classification of the party package and the partykit package showed that the female infants who received divorced or widowed women had the highest incidence of injury, followed by the uneducated women. Conclusion The application conditions of R software decision tree package are different, and should be based on the actual situation to select the appropriate software package for application analysis. In the classification of injury data modeling and analysis, the use of partykit package is more suitable, simple and user-friendly interface.