摘要
采用支持向量机(SVM)、随机森林(RF)和逻辑回归(LR)等机器学习方法对脑卒中患者进行分类研究,构建脑卒中疾病预测模型,以期为疾病发生提供早期预警。对kaggle网站下载healthcare-dataset-stroke-data的数据通过SMOTE智能过抽样算法构建均衡数据集,运用支持向量机、随机森林和逻辑回归算法构建脑卒中预测模型。将SMOTE算法优化前后的预测结果进行比较分析,并采用支持向量机、随机森林和逻辑回归算法对优化后的数据集构建疾病预测模型,其结果的准确率、精确度、召回率和ROC值都有明显提高。仿真实验结果可知SMOTE+随机森林算法预测模型的准确率、精确度、ROC值都优于支持向量机和逻辑回归预测模型,可用于脑卒中疾病的早期预测,为医疗手段干预赢得时间,对降低脑卒中的发病具有重要意义。
To study the classification of stroke patients by machine learning methods such as support vector machine(SVM),random forest(RF)and logistic regression(LR),and to construct a prediction model of stroke disease,so as to provide early warn⁃ing for disease occurrence.The data of health care dataset stroke data downloaded from kaggle website was used to construct the bal⁃anced data set by smote intelligent over sampling algorithm,and the stroke prediction model was constructed by using support vector machine,random forest and logistic regression algorithm.The prediction results of smote algorithm before and after optimization were compared and analyzed,and the disease prediction model was constructed by using support vector machine,random forest and logistic regression algorithm.The accuracy,precision,recall rate and ROC value of the results were significantly improved.The sim⁃ulation results show that the accuracy,accuracy and ROC value of smote+random forest algorithm prediction model are better than those of support vector machine and logistic regression prediction model,which can be used for the early prediction of stroke dis⁃ease,win time for medical intervention,and is of great significance to reduce the incidence of stroke.
作者
郭志恒
刘青萍
刘芳
王成武
阮旭凌
GUO Zhiheng;LIU Qingping;LIU Fang;WANG Chengwu;RUAN Xuling(Management and Information Engineering School,Hunan University of Chinses Medicine,Changsha 410208;The Domestic First-class Discipline Construction Project of Chinese Medicine,Changsha 410208;Pharmacy School,Hunan University of Chinese Medicine,Changsha 410208)
出处
《计算机与数字工程》
2021年第11期2180-2183,2247,共5页
Computer & Digital Engineering
基金
国家自然科学青年项目(编号:81704064)
2020年湖南省学位与研究生教育改革研究项目(编号:2020JGYB132)
湖南中医药大学中医学国内一流建设学科开放基金(编号:2018ZYX17)
湖南中医药大学信息科学与工程学院电子科学与技术学科开放基金(编号:2018-2)资助。
