面向糖尿病数据集挖掘频繁项目集和关联规则  

Mining frequent itemsets and association rules for diabetes datasets

作　　者：耿飙[1] 梁成全 Geng Biao;Liang Chengquan(Foundation Department,Suzhou Vocational Health College,Suzhou,Jiangsu 215009,China;Information Section,Huadong Sanatorium)

机构地区：[1]苏州卫生职业技术学院基础部,江苏苏州215009 [2]华东疗养院信息科

出　　处：《计算机时代》2022年第12期52-55,共4页Computer Era

基　　金：苏州卫生职业技术学院校级领雁培育重点项目(szwzy202004)。

摘　　要：为了从大型数据集中找到数据不同属性之间的关系,对糖尿病数据集通过FP-Growth算法及其变体提取关联规则,即CFP-Growth和ICFP-Growth。通过实验,三种算法FP-Growth、CFP-Growth和ICFP-Growth的准确率分别为51.30%,57%和60.5%。结果表明ICFP-Growth比CFP-Growth和FP-Growth更准确。In order to find the relationship between different attributes of data from large datasets,association rules are extracted from diabetes datasets through FP-Growth algorithm and its variants,namely CFP-Growth and ICFP-Growth.The accuracies of the three algorithms FP-Growth,CFP-Growth and ICFP-Growth are obtained through experiments as 51.30%,57%and 60.5%,respectively.The results show that ICFP-Growth is more accurate than CFP-Growth and FP-Growth.

关 键 词：数据挖掘 关联规则 频繁项目集 FP-GROWTH CFP-Growth ICFP-Growth 

分 类 号：TP399[自动化与计算机技术—计算机应用技术]