Binary Simulated Normal Distribution Optimizer for feature selection: Theory and application in COVID-19 datasets

作者：

Highlights：

• A hybrid version of GNDO with Simulated Annealing is proposed.

• SA is used as local search to achieve higher classification accuracy.

• The proposed method is evaluated on 18 well-known UCI datasets.

• The proposed method is tested on high dimensional microarray datasets.

• It is also applied on a COVID-19 dataset for the classification purpose.

摘要

•A hybrid version of GNDO with Simulated Annealing is proposed.•SA is used as local search to achieve higher classification accuracy.•The proposed method is evaluated on 18 well-known UCI datasets.•The proposed method is tested on high dimensional microarray datasets.•It is also applied on a COVID-19 dataset for the classification purpose.

论文关键词：Meta-heuristic,Feature selection,Generalized Normal Distribution Optimizer,Simulated annealing,COVID-19,Optimization,Algorithm

论文评审过程：Received 13 December 2020, Revised 25 February 2022, Accepted 3 March 2022, Available online 15 March 2022, Version of Record 29 March 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.116834