Risk prediction model of diabetic nephropathy based on machine learning
LIU Kui
HAN Zhengyuan
LI Linyi
MA Yifei
YANG Ziwei
ZHAO Luyang
CHEN Changsheng
WAN Yi
Abstract：Objective To analyze and compare the application of machine learning-based LightGBM and random forest model in the disease risk prediction model of diabetic nephropathy(DN)in diabetic patients.Methods Based on the public data set of National Population Health Data Center,the disease risk prediction model of DN was established by LightGBM and random forest algorithms for analysis and comparative study.Results The accuracy,area under the curve(AUC),recall rate,precision and F1 score of LightGBM prediction model were 0.775 0,0.807 1,0.596 3,0.878 0,and 0.7102,respectively,which were all higher than those of random forest model.Conclusion Machine learning-based LightGBM and random forest model have good prediction effect on DN.LightGBM has higher accuracy,AUC,precision,recall rate and F1 score.In clinical practice,machine learning can be used to provide reference for related researches.
Keywords：diabetic nephropathymachine learningprediction analysiscomparative study
Publication Date：2025-02-27
Online Publishing Date：2026-08-26(First online date of this platform, not the publication date of the document)
Pages：6( 226-231 )