基于机器学习的宫颈癌致病因素分析  

作　　者：潘凤 王杰 张艳莎 王林 

机构地区：[1]贵州民族大学数据科学与信息工程学院,贵州贵阳550025 [2]贵州省模式识别与智能系统重点实验室,贵州贵阳550025

出　　处：《工业控制计算机》2023年第4期122-124,共3页Industrial Control Computer

摘　　要：宫颈癌是损害女性健康的疾病之一,其致病与个人的生活习惯有着重要关系。基于UCl中的cervical cancer(risk factors)数据集,采用Boruta算法筛选特征,并利用XG-Boost算法建立宫颈癌致病因素模型。分别使用Hinselmann、Schiller、Citology和Biopsy四种检测方法得到的预测精度依次为91%、90%、93.3%和87%。实验结果显示:致病因素中不良生活习惯占据多数,表明女性的不良生活习惯对女性患宫颈癌具有显著影响。Cervical cancer is one of the diseases that harm women's health,and its pathogenesis has an important relationship with personal living habits.Therefore,based on the cervical cancer(risk factors)datasets in UCl,Boruta algorithm is used to screen the features,and XG-Boost algorithm is used to establish the cervical cancer pathogenesis factor model.The four detection methods,prediction accuracies obtained using Hinselmann,Schiller,Citology,and Biopsy are successively 91%,90%,93.3%,and 87%,respectively.The experimental results showed that the bad habits accounted for the majority of the pathogenic factors,indicating that women's bad habits had a significant impact on women's cervical cancer.

关 键 词：宫颈癌预测 致病因素 生活习惯 Boruta算法 XG-Boost算法 

分 类 号：R737.33[医药卫生—肿瘤] TP181[医药卫生—临床医学]