摘要
目的使用关联规则的数据挖掘方法研究一些慢性病之间的关联性及关联强度。方法将调查的4858名社区居民的患病情况作为数据集,使用SPSS Clementine 12.0软件中的Apriori modeling算法对数据集进行分析。结果在4585名调查人群中,患病率排在前三位的疾病分别为高血压(16.51%)、高血脂(11.20%)、痛风及骨关节病(10.44%)。关联规则分析的结果为,在筛选出的10条强关联中,按置信度大小排在前三位的分别为高血脂→高血压,置信度为50.18%、规则支持度为5.62%、提升比为3.04;糖尿病→高血压,置信度为48.81%、规则支持度为2.94%、提升比为2.96;脂肪肝等肝脏疾病→高血脂,置信度为40.98%、规则支持度为2.76%、提升比为3.66。结论高血压、糖尿病和高血脂不仅三者之间存在关联,而且与其他疾病也存在广泛的关联,提示在患有这三种疾病的情况下应注意与之相关联的疾病的筛查和预防。
Objective The purpose of the study was to ana-lyze the associations and association intension among some chronic diseases using association role mining （ ARM ）. Methods The suffering condi- tions of 4858 residents were constructed a data mart for analysis. Apriori modeling of the ARM method in the Clementine 12. 0 program were used to analyze the data mart. Results In 4858 residents, the diseases of the top three prevalence were hypertension （16.51%）, hyperlipidemia （ 11.20% ） ,gout and osteoarthropathy（ 10. 44% ）. As a result of ARM, in the selected 10 rules, the rule with the highest confidence was hyperlipi- demia to hypertension, confidence was 50. 18 %, rules support was 5.62%, lift was 3.04. The second rule was diabetes to hypertension, confidence was 48. 81% ,rules support was 2.94% ,lift was 2.96. The third rule was Fatty liver and liver disease to hyperlipidemia, confidence was 40.98%, rules support was 2.76% ,lift was 3.66. Conclusion There were association rules among hypertension, diabetes and hyperlipidemia. This three diseases also had associations with other diseases. So, when a patient had the three diseases ,he or she should pay attention to screening and prevent the disea- ses that had association with the three diseases.
出处
《中国卫生统计》
CSCD
北大核心
2013年第6期818-820,共3页
Chinese Journal of Health Statistics
关键词
关联规则
慢性病
数据挖掘
患病率
Association rule
Chronic disease
Data min- ing
Prevalence
