Comparison of NN and LR classifiers in the context of screening native American elders with diabetes
作者：
Highlights：
• By year 2030 1 out of every 3 American will be diabetic, this is even worse if we consider AI/AN elders.
• Most of the members of this community reside in rural areas where healthcare resources are sparse.
• Data mining tools are an accepted alternative to early screening and preventive intervention.
• LR and ANN are two very popular data mining tools that are applied to tackle wide variety of issues.
• In this context both LR and NN have similar classification ability however NN was marginally better.
摘要
•By year 2030 1 out of every 3 American will be diabetic, this is even worse if we consider AI/AN elders.•Most of the members of this community reside in rural areas where healthcare resources are sparse.•Data mining tools are an accepted alternative to early screening and preventive intervention.•LR and ANN are two very popular data mining tools that are applied to tackle wide variety of issues.•In this context both LR and NN have similar classification ability however NN was marginally better.
论文关键词：Artificial Neural Network,Logistic Regression,Genetic algorithm,Stepwise regression,Diabetes,AI/AN elders
论文评审过程：Available online 16 May 2013.
论文官网地址：https://doi.org/10.1016/j.eswa.2013.05.012
