摘要
针对目前透析时机选择不准确的问题,提出基于优化集成学习的预测模型以提高透析时机预测精度。在模型应用方面,根据十种不同分类器的对比实验结果,利用RF、AdaBoost、GBDT等集成学习器和SVM、KNN、NB、DT、MLP、ET、LR等分类器进行模型融合以降低误差。实验结果表明采用优化集成学习的模型拥有更好的泛化能力(95.04%)和ROC曲线(97%)以及更低的标准误差(+/-0.04)。
Aiming at the problem of inaccurate hemodialysis timing,proposes a model based on ensemble learning for improving prediction accuracy of hemodialysis timing.In the model application,according to the comparison of experimental results of the ten classifier models,utilizes in.tegrated learners such as RF,AdaBoost,GBDT and classifiers such as SVM,KNN,NB,DT,MLP,ET,LR for model fusion to reduce errors.The experimental results show that the ensemble learning model has better generalization ability(97.04%)and ROC(99%)and lower stan.dard error(+/-0.04).
作者
熊长柱
李健
XIONG Chang-zhu;LI Jian(College of Electronics and Information Engineering,Sichuan University,Chengdu 610065)
出处
《现代计算机》
2019年第9期14-17,共4页
Modern Computer
关键词
集成学习
预测
机器学习
分类
Ensemble Learning
Prediction
Machine Learning
Classification
