Hadoop平台下基于朴素贝叶斯算法的心脏疾病预测方案  

Heart disease prediction solution based on Naive Bayes algorithm under Hadoop platform

作　　者：王自强 尚志会 石永华 王毅 符萱 杨生正 Wang Ziqiang;Shang Zhihui;Shi Yonghua;Wang Yi;Fu Xuan;Yang Shengzheng(College of Medical Information Engineering,Zunyi Medical University,Zunyi 563000,China)

机构地区：[1]遵义医科大学医学信息工程学院,遵义563000

出　　处：《现代计算机》2024年第11期16-22,共7页Modern Computer

基　　金：贵州省教育厅高校人文社会科学研究项目(23RWJD162);贵州省卫生健康委科学技术基金项目(gzwkj2022-524);大学生创新创业训练计划项目(ZYDC202301099);贵州省科技计划项目(黔科平台人才[2020]-030);贵州省高等学校教学内容与课程体系改革项目(SJJG2022-02-172);遵义市科技计划项目(遵市科合HZ字[2023]191号);遵义医科大学2021年度学术新苗培养及创新探索专项项目(黔科平台人才[2021]1350-027)。

摘　　要：应用分布式存储、计算技术和朴素贝叶斯算法,构建心脏病预测模型方案。首先搭建Hadoop完全分布式平台,结合Python语言和MapReduce编程框架构建朴素贝叶斯分类器。此外,用MapReduce实现对算法并行化以提升分析效率。以2020美国CDC数据集作为心脏病数据集,在测试集中算法准确率达到88.52%,且经验证该方案能够在实际使用中成功预测是否患心脏病。此方案准确率较高且具备高可扩展性、分布式存储和计算、容错性等优势,形成了一种可靠、高效和低成本的解决方案。We have developed a heart disease prediction model using distributed storage,computing technology,and the naive Bayes algorithm.Firstly,we built a Hadoop fully distributed platform and combined it with the Python language and MapReduce programming framework to construct the naive Bayes classifier.Additionally,we parallelized the algorithm using MapReduce to im-prove analysis efficiency.Using the 2020 US CDC dataset as the heart disease dataset,the accuracy of the algorithm reached 88.52% in the test set.Furthermore,our solution has been validated to be able to successfully predict whether an individual is suffering from heart disease in practical applications.This solution has high accuracy and advantages such as high scalability,distributed storage and computing,and fault tolerance,forming a reliable,efficient,and low-cost solution.

关 键 词：HADOOP MAPREDUCE 数据挖掘 朴素贝叶斯 

分 类 号：R541[医药卫生—心血管疾病] TP18[医药卫生—内科学]