A new nearest neighbor-based framework for diabetes detection

作者：

Highlights：

• A new nearest neighbor-based framework is proposed to classify diabetic patients.

• It consists of k-means clustering, autoencoder, and nearest neighbor classifier.

• A new variant of the nearest neighbor-based classifier is also proposed.

• Evaluations are performed on binary-class and multiclass diabetes datasets.

• This new framework is better than the recent deep learning-based model.

摘要

•A new nearest neighbor-based framework is proposed to classify diabetic patients.•It consists of k-means clustering, autoencoder, and nearest neighbor classifier.•A new variant of the nearest neighbor-based classifier is also proposed.•Evaluations are performed on binary-class and multiclass diabetes datasets.•This new framework is better than the recent deep learning-based model.

论文关键词：Diabetes detection,k-nearest neighbor,k-means clustering,Dimensional reduction,Multi-voter multi-commission

论文评审过程：Received 30 April 2021, Revised 27 November 2021, Accepted 7 March 2022, Available online 23 March 2022, Version of Record 29 March 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.116857