摘要
为帮助有效识别患有妊娠期糖尿病的孕妇,提前干预治疗,降低其潜在风险,设计并开发了妊娠期糖尿病智能预测系统。首先,在对比10种机器学习模型的基础上对临床数据集进行数据归一化和特征筛选,减少模型计算复杂度和一些使模型不稳定的特征;其次,使用Stacking算法对10种不同的机器学习模型进行集成,分别搭建2个集成模型Stacking1和Stacking2,比较集成模型的性能;最后,基于PyQt5设计妊娠期糖尿病智能预测系统,预测孕妇患妊娠期糖尿病的风险,并给出建议。结果表明:使用10个机器学习模型以及2个集成模型对GDM进行预测,发现GBDT的预测结果高于其他机器学习模型,集成模型Stacking2将多个异质学习器相结合表现出较高的准确性和可靠性,且评价指标Accuracy、Precision、Recall、AUC分别为0.900 9、0.901 2、0.900 7、0.900 7,均高于同类模型。智能预测系统能有效预测妊娠期糖尿病的风险,能够及早发现易患病人群,并提供妊娠期糖尿病的科普知识,从而加强对易患病人群的健康管理,降低妊娠期糖尿病发生的风险。
To help to effectively identify pregnant women with gestational diabetes mellitus(GDM),intervene in advance for treatment,and reduce potential risks,a GDM intelligent prediction system was designed and developed.Firstly,based on the comparison of 10 machine learning models,data normalization and feature selection were conducted on the clinical dataset based on these models to reduce model computational complexity and unstable features.Secondly,the Stacking algorithm was used to integrate 10 different machine learning models,and two integrated models,Stacking1 and Stacking2,were built to compar their performance.Finally,based on PyQt5,a GDM intelligent prediction system was designed to predict the risk of GDM in pregnant women and provide suggestions.The results show that using 10 machine learning models and two integrated models to predict GDM,it is found that the prediction results of GBDT are higher than other machine learning models.The integrated model Stacking2,combining multiple heterogeneous learners,exhibits high accuracy and reliability,with evaluation indicators of Accuracy,Precision,Recall,and AUC of 0.9009,0.9012,0.9007,and 0.9007,respectively,all higher than similar models.The intelligent prediction system can effectively predict the risk of GDM,identify susceptible individuals early,and provide popular science knowledge of GDM,thereby strengthen the health management of susceptible individuals and reduce the risk of GDM occurrence.
作者
马金龙
徐立魏
杨志芬
王胜普
郑芮
张永强
MA Jinlong;XU Liwei;YANG Zhifen;WANG Shengpu;ZHENG Rui;ZHANG Yongqiang(School of Information Science and Engineering,Hebei University of Science and Technology,Shijiazhuang,Hebei 050018,China;Department of Obstertrics,The Fourth Hospital of Hebei Medical University,Shijiazhuang,Hebei 050035,China)
出处
《河北工业科技》
CAS
2024年第2期99-107,共9页
Hebei Journal of Industrial Science and Technology
基金
河北省自然科学基金(H2022206212)
河北省医学科学研究课题计划项目(20230775,20230776,20210715)。
关键词
计算机决策支持系统
机器学习
PyQt5
集成学习
妊娠期糖尿病
特征筛选
computer decision support system
machine learning
PyQt5
integrated learning
gestational diabetes mellitus
feature selection
